Spoilers For Monday January 26th 2015

[James Robinson]

COLIN yet again was fantastic, and now we return to normality, as amazingly, January is almost done with already

So, we're in mid-December, just over a week before Christmas. What does a guy do to entertain himself He goes to MediaCityUK to watch his favourite show of course.

Even more amazingly, I find out that Tracey Mills is the reigning champion. (YAY!!) (OK, I had an idea she was the champ, but I didn't know what had happened at all, apart from tiny snippets from Kirk.) So, I was there hoping for the best for Tracey, maybe a bit of extra good luck would be with her from yours truly.

She's up for half-octochampdom today, and Mickie Wynne-Davies is the only contestant standing in her way of that important milestone.

In for her 3rd spell in DC this week is ex-Blue Peter presenter, Helen Skelton.

Join Andy for the recap later.

[Tony Atkins]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

[Jennifer Steadman]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

[Tony Atkins]

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

Apterous allows it.

Alt to ZENITHS in R10 of HISTONE.

[Jennifer Steadman]

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

Apterous allows it.

Alt to ZENITHS in R10 of HISTONE.

Apterous allows a number of words that (rightly or wrongly) would never be accepted on TV, so that's not really much help.

[Jennifer Steadman]

The challenger was brilliant TV, her conundrum buzz properly cracked me up

[Tony Atkins]

No standard method available for the 3L game (I did Rachel's way).

[Tony Atkins]

The challenger was brilliant TV, her conundrum buzz properly cracked me up

I enjoyed her comments during the selections.

[Keith Bennett]

No standard method available for the 3L game (I did Rachel's way).

(3x100)-(9x4)+(75/25)

Don't know if you'd call it "standard" but it worked well enough for me.

[Gavin Chipper]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

Weirdly the search thing can't seem to find it. I think WAISTED has traditionally been allowed and that's the same. It was last offered by DC on 20th November 2014.

[Tracey Anne Mills]

I had FACETED but didn't risk it

[Graeme Cole]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

I've always understood the confusingly-named "in combination" rule to mean that if a word appears as part of a multi-word entry (e.g. MISTLE in MISTLE THRUSH), that's not enough to allow it - it has to appear on its own. I don't think it's anything to do with the "[in combination]" tag. There are lots of words that don't make sense unless they appear with other words (e.g. THAN), but they're perfectly allowable. What's different with words like MISTLE is that they only appear in combination with specific words.

SEATER has been disallowed before, because its entry in the dictionary is "-seater", but FACETED doesn't have a hyphen before it.

Weirdly the search thing can't seem to find it. I think WAISTED has traditionally been allowed and that's the same. It was last offered by DC on 20th November 2014.

There are some valid words which the search box won't find by default. At COLIN on Saturday I noticed that you couldn't find MANNERIST in the dictionary unless you happened to look at the entry for MANNERISM. If you search for MANNERIST it takes you to a Spanish translation. You have to change the drop-down box to "British and World English" to make MANNERIST take you to its entry under MANNERISM.

Fiddling with the drop-down box still doesn't make it recognise FACETED, though - it takes you to a paywalled entry in the New Oxford Dictionary for Writers and Editors.

[Gavin Chipper]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

I've always understood the confusingly-named "in combination" rule to mean that if a word appears as part of a multi-word entry (e.g. MISTLE in MISTLE THRUSH), that's not enough to allow it - it has to appear on its own. I don't think it's anything to do with the "[in combination]" tag. There are lots of words that don't make sense unless they appear with other words (e.g. THAN), but they're perfectly allowable. What's different with words like MISTLE is that they only appear in combination with specific words.

SEATER has been disallowed before, because its entry in the dictionary is "-seater", but FACETED doesn't have a hyphen before it.

Weirdly the search thing can't seem to find it. I think WAISTED has traditionally been allowed and that's the same. It was last offered by DC on 20th November 2014.

There are some valid words which the search box won't find by default. At COLIN on Saturday I noticed that you couldn't find MANNERIST in the dictionary unless you happened to look at the entry for MANNERISM. If you search for MANNERIST it takes you to a Spanish translation. You have to change the drop-down box to "British and World English" to make MANNERIST take you to its entry under MANNERISM.

Fiddling with the drop-down box still doesn't make it recognise FACETED, though - it takes you to a paywalled entry in the New Oxford Dictionary for Writers and Editors.

Thanks for the information.

[Adam Gillard]

The challenger was brilliant TV, her conundrum buzz properly cracked me up

I enjoyed her comments during the selections.

Agreed. Mickie was amazing!

[James Robinson]

Yeah, it was a great show to be in the audience for to start the day with.

The only "complaint" I have is that they cut out quite a massive chunk of Nick, Rachel and DC talking about where BIGAMOUS/TRIGAMOUS/MONOGAMOUS were discussed in great detail.

[Adam Dexter]

I had a word towards the end that doesn't seem to be valid but I thought it was a word - UMBONGI(S?). What word am I thinking of?

[Innis Carson]

Probably IMBONGI, and perhaps Um Bongo.

[Jon O'Neill]

A few alts about so far (eg OCTUPLE+T), but just one 7 beater available for R8 (DEFEAT+COM).

Not sure if the 7 beater would be valid - it's listed as a derivative of FACET, but has a note to say it can only be used 'in combination', so it's really more like -FACETED. Is there anything in the rules that specifies whether combination words are valid or not?

I've always understood the confusingly-named "in combination" rule to mean that if a word appears as part of a multi-word entry (e.g. MISTLE in MISTLE THRUSH), that's not enough to allow it - it has to appear on its own. I don't think it's anything to do with the "[in combination]" tag. There are lots of words that don't make sense unless they appear with other words (e.g. THAN), but they're perfectly allowable. What's different with words like MISTLE is that they only appear in combination with specific words.

SEATER has been disallowed before, because its entry in the dictionary is "-seater", but FACETED doesn't have a hyphen before it.

Weirdly the search thing can't seem to find it. I think WAISTED has traditionally been allowed and that's the same. It was last offered by DC on 20th November 2014.

There are some valid words which the search box won't find by default. At COLIN on Saturday I noticed that you couldn't find MANNERIST in the dictionary unless you happened to look at the entry for MANNERISM. If you search for MANNERIST it takes you to a Spanish translation. You have to change the drop-down box to "British and World English" to make MANNERIST take you to its entry under MANNERISM.

Fiddling with the drop-down box still doesn't make it recognise FACETED, though - it takes you to a paywalled entry in the New Oxford Dictionary for Writers and Editors.

I don't get that odd outcome when I search for MANNERIST, but I have seen some weird stuff before, so I believe that the behaviour could've changed since you wrote your post.

Another example: searching for MANKINDS leads you to an article about Sentence Length..?

This all convincingly undermines Gevin's Search Box Word Validation method, consequently necessitating the presence of an expert adjudicator like Susie.

[Adam Dexter]

Probably IMBONGI, and perhaps Um Bongo.

IMBONGI - thanks Can I stick an S on it?

[Innis Carson]

Probably IMBONGI, and perhaps Um Bongo.

IMBONGI - thanks Can I stick an S on it?

Yep, IMBONGIS is specified as a plural (along with IIMBONGI and IZIMBONGI).

[Gavin Chipper]

This all convincingly undermines Gevin's Search Box Word Validation method, consequently necessitating the presence of an expert adjudicator like Susie.

Oh, so you think that because Graeme has posted you can come along and hide behind him rather than just whine to me by text like you normally do whenever you disagree with anything I post.

But anyway, it seems you're taking an all-or-nothing approach. I originally suggested the search box method when it was first decided we were using the online dictionary, and was open to any improvements based on any problems with it. It's just that you never bothered to come up with anything. Actually a few things have come up anyway, such as whether to allow FLORUITED based on the fact that it isn't found, even though it's in an example sentence and it's the logical past tense. I've acknowledged problems before anyway, which you'd know if you'd been paying attention.

I think the best approach would still involve using the search box. Particularly with the plurals of mass nouns, I'd use it as the arbiter. Generally, I'd say a word being found using that is a sufficient condition for allowing a word, but not a necessary one.

But as a good approximation to the best method, I'd say everything found using the search box and everything explicitly listed should be allowed. So we'd obviously allowed FACETED. We still could get cases where the derivative is very different from the headword, so it's not obvious where to find it (such as when Kirk got HEDARIM), but it would require both that and the search thing not finding it, which should be very rare.

But regardless of exactly the best method is, it's not "Throw away any idea of getting proper formalised rules and let Susie make it up as she goes along".

[Jon O'Neill]

Of course you're being ridiculous. I haven't suggested any alternative, because YOUR approach is the alternative to the status quo. And it is not an improvement.

Just to restate, the status quo is to apply the rules of Countdown (as they have more or less stood for 10+ years) to the headwords found in ODO. These rules have been misapplied in the past but that doesn't mean we should dismiss it as the best solution.

Your solution is poor because it relies on a database table of HTTP 302 redirects, which is opaque, could change at the whim of a web developer, and has been shown to be completely bananas (pattys, patties, audios and mankinds all link to pages in Garner's Legal Dictionary..).

This is not to mention that it can actually be shown to lead to completely wrong results. If I spent an hour I could probably find loads of examples of stuff like this, but the best example from the 5 minutes I did spend is to look up ABOUTS, which redirects to GADABOUT. Is ABOUTS a valid plural of GADABOUT now? Do we accept it just because?

[Gavin Chipper]

Of course you're being ridiculous. I haven't suggested any alternative, because YOUR approach is the alternative to the status quo. And it is not an improvement.

Just to restate, the status quo is to apply the rules of Countdown (as they have more or less stood for 10+ years) to the headwords found in ODO. These rules have been misapplied in the past but that doesn't mean we should dismiss it as the best solution.

Well, when Damian first announced that the online dictionary would be used, my understanding was that it was completely open what was going to happen, and that's why they delayed its implementation by a series - i.e. there effectively was no status quo, and the rules were going to be looked at afresh.

And why do you think the rules have so frequently been misapplied? There are various possible answers, but the most generous one is that the rules aren't precise enough and they're not actually the best solution just because they've been around for 10+ years. Or if they are the best, we're effectively saying that there is no possible good set of rules. I also, perhaps naively, thought that one of the reasons for the implementation of the online dictionary was to make things more objective. So when, after delaying it by a series, we were told that nothing had changed, I found it a bit disappointing.

Your solution is poor because it relies on a database table of HTTP 302 redirects, which is opaque, could change at the whim of a web developer, and has been shown to be completely bananas (pattys, patties, audios and mankinds all link to pages in Garner's Legal Dictionary..).

It didn't redirect me when I searched for PATTIES. But yeah, I understand it's not perfect, but that doesn't mean you can't define rules around it. For a start, no-one gives a shit about terms like "HTTP 302 redirects" - trying to bamboozle us with obscure terms doesn't change anything. And while it is opaque, in that the relevant information isn't publicly viewable, I don't see why it's any more at the whim of a web developer than anything else. Presumably decisions have been made about which mass nouns can be pluralised, and that's why some are found and some aren't. I think we have to take it on trust that web developers aren't going to go rogue and start trolling the system.

This is not to mention that it can actually be shown to lead to completely wrong results. If I spent an hour I could probably find loads of examples of stuff like this, but the best example from the 5 minutes I did spend is to look up ABOUTS, which redirects to GADABOUT. Is ABOUTS a valid plural of GADABOUT now? Do we accept it just because?

No. We don't. I gave an approximation to the rules we should use, and these can be honed further. Another addition to this could be that if you search for a word and it finds an entry where it would make no logical sense for it to be a derivative of the headword and it's not mentioned as a derivative (and yes, this could easily be defined more precisely), then we wouldn't allow it purely on that basis.

[Matt Morrison]

C4C is alive again.

[Jon O'Neill]

Of course you're being ridiculous. I haven't suggested any alternative, because YOUR approach is the alternative to the status quo. And it is not an improvement.

Just to restate, the status quo is to apply the rules of Countdown (as they have more or less stood for 10+ years) to the headwords found in ODO. These rules have been misapplied in the past but that doesn't mean we should dismiss it as the best solution.

Well, when Damian first announced that the online dictionary would be used, my understanding was that it was completely open what was going to happen, and that's why they delayed its implementation by a series - i.e. there effectively was no status quo, and the rules were going to be looked at afresh.

Well you're either wrong on purpose because you're a mad bastard or you're wrong because you're wrong:

Sorry, bad example then. TRUANCIES is a better example. What I'm getting at is, if a player offers TRUANCIES for 9, is the process: look up "truancy" on ODO, see that it's a mass noun, and disallow it; or look up "truancies", see that it gives you the entry for "truancy" (which it does) rather than saying it's unrecognised like it does for some other mass noun plurals, so allow it? Or something else?

We look up the word offered, so would enter TRUANCIES. If no results are shown in the search then it would be disallowed, but in this case, we'd analyse the entry for any evidence that points to accepting it.

"Analyse the entry for evidence that points to accepting it" sounds suspiciously like the current rules. Further evidence:

Great move. Regarding Graeme's point, I'd think that if the search bar recognises a plural and redirects you to the root entry, then that should be taken as an unequivocal indication that the plural is valid. If it doesn't recognise the plural, then you would have to look at the entry and make the decision in the 'traditional' way.

Yes i think that's more or less the method we'll use, but we're still getting directions from OUP about the ODO website and how it offers up results etc, but by and large i don't think you're far off with what you've said.

So just to be clear, the rules were never changing. You invented that.

And why do you think the rules have so frequently been misapplied? There are various possible answers, but the most generous one is that the rules aren't precise enough and they're not actually the best solution just because they've been around for 10+ years. Or if they are the best, we're effectively saying that there is no possible good set of rules. I also, perhaps naively, thought that one of the reasons for the implementation of the online dictionary was to make things more objective. So when, after delaying it by a series, we were told that nothing had changed, I found it a bit disappointing.

Is one of the "various possible answers" that the world is out to get you, led by Damian Eadie and Susie Dent?

Your solution is poor because it relies on a database table of HTTP 302 redirects, which is opaque, could change at the whim of a web developer, and has been shown to be completely bananas (pattys, patties, audios and mankinds all link to pages in Garner's Legal Dictionary..).

It didn't redirect me when I searched for PATTIES. But yeah, I understand it's not perfect, but that doesn't mean you can't define rules around it. For a start, no-one gives a shit about terms like "HTTP 302 redirects" - trying to bamboozle us with obscure terms doesn't change anything. And while it is opaque, in that the relevant information isn't publicly viewable, I don't see why it's any more at the whim of a web developer than anything else. Presumably decisions have been made about which mass nouns can be pluralised, and that's why some are found and some aren't. I think we have to take it on trust that web developers aren't going to go rogue and start trolling the system.

I meant PATTIE, apologies.

For a start, there might be some people who are interested in the mechanics behind the idea you are suggesting. I naively thought you might be one of those people. But never mind. What we know is that there is information somewhere which tells us what search terms redirect to what page. What we don't know is who decided what redirects where, and whether they are applying the guidance notes from the front of the ODE any better than Susie does. What we also don't know is whether they are even trying to apply those notes. What we also don't know is why some borderline entries redirect to a Legal Style Guide. What we also don't know is why some borderline entries redirect to completely different words.

This is not to mention that it can actually be shown to lead to completely wrong results. If I spent an hour I could probably find loads of examples of stuff like this, but the best example from the 5 minutes I did spend is to look up ABOUTS, which redirects to GADABOUT. Is ABOUTS a valid plural of GADABOUT now? Do we accept it just because?

No. We don't. I gave an approximation to the rules we should use, and these can be honed further. Another addition to this could be that if you search for a word and it finds an entry where it would make no logical sense for it to be a derivative of the headword and it's not mentioned as a derivative (and yes, this could easily be defined more precisely), then we wouldn't allow it purely on that basis.

So, in your proposed method, there is scope for the goalposts to be moved if the results don't match what we'd expect? And how is this supposed to be an improvement again? Not only does your method not work as well for a TV Gameshow, it also doesn't work as well for a rigidly-defined game. It's time to give up until we have some more from OUP.

[George Pryn]

I meant PATTIE, apologies.

[Gavin Chipper]

Of course you're being ridiculous. I haven't suggested any alternative, because YOUR approach is the alternative to the status quo. And it is not an improvement.

Just to restate, the status quo is to apply the rules of Countdown (as they have more or less stood for 10+ years) to the headwords found in ODO. These rules have been misapplied in the past but that doesn't mean we should dismiss it as the best solution.

Well, when Damian first announced that the online dictionary would be used, my understanding was that it was completely open what was going to happen, and that's why they delayed its implementation by a series - i.e. there effectively was no status quo, and the rules were going to be looked at afresh.

Well you're either wrong on purpose because you're a mad bastard or you're wrong because you're wrong:

Sorry, bad example then. TRUANCIES is a better example. What I'm getting at is, if a player offers TRUANCIES for 9, is the process: look up "truancy" on ODO, see that it's a mass noun, and disallow it; or look up "truancies", see that it gives you the entry for "truancy" (which it does) rather than saying it's unrecognised like it does for some other mass noun plurals, so allow it? Or something else?

We look up the word offered, so would enter TRUANCIES. If no results are shown in the search then it would be disallowed, but in this case, we'd analyse the entry for any evidence that points to accepting it.

"Analyse the entry for evidence that points to accepting it" sounds suspiciously like the current rules. Further evidence:

Great move. Regarding Graeme's point, I'd think that if the search bar recognises a plural and redirects you to the root entry, then that should be taken as an unequivocal indication that the plural is valid. If it doesn't recognise the plural, then you would have to look at the entry and make the decision in the 'traditional' way.

Yes i think that's more or less the method we'll use, but we're still getting directions from OUP about the ODO website and how it offers up results etc, but by and large i don't think you're far off with what you've said.

So just to be clear, the rules were never changing. You invented that.

Maybe I was overly optimistic about the rules being looked at completely afresh, but you are also wrong. Innis suggested that if a plural of a mass noun is found then it should be automatically allowed, and Damian appeared to agree with that. That is a change. And the next post after as well:

It's been noticed that ODO will recognise some comparatives and superlatives of monosyllabic adjectives (from what I've seen, mainly everyday ones like WILDEST and GRANDER) and redirect you to the root word, whereas other ones (such as apterous abominations FAUXER and HINGEDEST) throw up "no results found". Will the show now take these as official adjudications of whether or not these words are valid, or carry on allowing all comparatives/superlatives of single syllable adjectives?

From memory (am not at work today and don't have a dictionary at home), the ODE3 mentions 1-syllable adjectives and the comparatives and superlatives in the notes section at the front, and i think it says 'in most cases' but not in every case. We'll get this checked out though because it's a great point you make and it does need clarifying for definite. We're not running with this until late November so there are still things that need ironing out. Cheers, Innis.

Both the two points in response to Innis suggest that they were going to take advice from OUP about what the rules should be rather than just leave things as they are. That suggests that things were open. The fact that they ultimately changed nothing isn't a vindication of what you've said.

And why do you think the rules have so frequently been misapplied? There are various possible answers, but the most generous one is that the rules aren't precise enough and they're not actually the best solution just because they've been around for 10+ years. Or if they are the best, we're effectively saying that there is no possible good set of rules. I also, perhaps naively, thought that one of the reasons for the implementation of the online dictionary was to make things more objective. So when, after delaying it by a series, we were told that nothing had changed, I found it a bit disappointing.

Is one of the "various possible answers" that the world is out to get you, led by Damian Eadie and Susie Dent?

No. I know you think otherwise, but a lot of people would be much happier with a clear set of rules for determining the validity of words, and find the current situation unsatisfactory.

Your solution is poor because it relies on a database table of HTTP 302 redirects, which is opaque, could change at the whim of a web developer, and has been shown to be completely bananas (pattys, patties, audios and mankinds all link to pages in Garner's Legal Dictionary..).

It didn't redirect me when I searched for PATTIES. But yeah, I understand it's not perfect, but that doesn't mean you can't define rules around it. For a start, no-one gives a shit about terms like "HTTP 302 redirects" - trying to bamboozle us with obscure terms doesn't change anything. And while it is opaque, in that the relevant information isn't publicly viewable, I don't see why it's any more at the whim of a web developer than anything else. Presumably decisions have been made about which mass nouns can be pluralised, and that's why some are found and some aren't. I think we have to take it on trust that web developers aren't going to go rogue and start trolling the system.

I meant PATTIE, apologies.

For a start, there might be some people who are interested in the mechanics behind the idea you are suggesting. I naively thought you might be one of those people. But never mind. What we know is that there is information somewhere which tells us what search terms redirect to what page. What we don't know is who decided what redirects where, and whether they are applying the guidance notes from the front of the ODE any better than Susie does. What we also don't know is whether they are even trying to apply those notes. What we also don't know is why some borderline entries redirect to a Legal Style Guide. What we also don't know is why some borderline entries redirect to completely different words.

Obviously I would prefer a more transparent system and would be interested in the mechanics of it, but throwing around jargonistic terms that most people on here wouldn't understand doesn't further the discussion. And while we don't know exactly what's going on, at least what is there is there and can be used to base rules around - rules that would mean we could make an unequivocal adjudication on a word, or at least much more frequently than is the case now. And even you think the results would be less "correct" than Susie's decisions, and made by some shadowy conspiratorial web developer, at least they'd be more consistent.

This is not to mention that it can actually be shown to lead to completely wrong results. If I spent an hour I could probably find loads of examples of stuff like this, but the best example from the 5 minutes I did spend is to look up ABOUTS, which redirects to GADABOUT. Is ABOUTS a valid plural of GADABOUT now? Do we accept it just because?

No. We don't. I gave an approximation to the rules we should use, and these can be honed further. Another addition to this could be that if you search for a word and it finds an entry where it would make no logical sense for it to be a derivative of the headword and it's not mentioned as a derivative (and yes, this could easily be defined more precisely), then we wouldn't allow it purely on that basis.

So, in your proposed method, there is scope for the goalposts to be moved if the results don't match what we'd expect? And how is this supposed to be an improvement again? Not only does your method not work as well for a TV Gameshow, it also doesn't work as well for a rigidly-defined game. It's time to give up until we have some more from OUP.

Who said anything about moving the goalposts? All of this rule honing could have been done before it was implemented. In the original thread, I said that it would be good if we could just use the search box, but let's look for examples where it might mess up. Then we could base the rules around everything we've discovered about it.

But, yes, we do need more from OUP. And what happened seems strange. Damian said they were clarifying stuff with them, but in the end nothing changed. I'm not sure what went wrong, but something did. I might send them another e-mail. Last time I did that, shortly afterwards they changed from the paper dictionaries to the ODO. And while I can't exactly claim credit, I would argue that there is a reasonable non-zero chance that it had some effect in terms of them getting their arses in gear. And if I do send another e-mail, as before I will post it in on the forum, so other people can copy it if they want to put the pressure on.

[Zarte Siempre]

Am I the only one thinking that Jono and Gev must have the most amazing angry make-up sex?

[Jon O'Neill]

The reason I didn't have this debate with you before is that a lot of words are wasted on what is quite a simple problem. I like to argue with you so I entertained myself a little with talk of who said what and why, but it's completely irrelevant, and as I suspected, has given credence to your logically unsound thinking.

We are advocating two different systems. The drawbacks of the system I am advocating:

1. Mass noun/other rules are not applied properly
2. Mass noun/other rules are not applied consistently in individual cases, leading to uncertainty

Does your system fix either of these?

1. No. I have no idea who decided what is or isn't a valid inflection/comparative/superlative/plural form in the ODO, but it doesn't take long to find some very objectionable results. See ABOUTS; also SALMONS which is redirected despite the plural being specified as SALMON.
2. Who knows? Since the list of redirects is completely opaque, we wouldn't know if it did change on an individual basis. When ABOUTS and SALMONS get fixed, how do we know? There is no extra certainty.

So your system doesn't fix any of the current issues. So does your system have any drawbacks?

1. Opaque - things could be allowed or disallowed without us knowing why.

You've already agreed with this so I won't go on making other points.. they're unnecessary to prove that my system is better.

[Jon O'Neill]

For a start, there might be some people who are interested in the mechanics behind the idea you are suggesting. I naively thought you might be one of those people. But never mind. What we know is that there is information somewhere which tells us what search terms redirect to what page. What we don't know is who decided what redirects where, and whether they are applying the guidance notes from the front of the ODE any better than Susie does. What we also don't know is whether they are even trying to apply those notes. What we also don't know is why some borderline entries redirect to a Legal Style Guide. What we also don't know is why some borderline entries redirect to completely different words.

Obviously I would prefer a more transparent system and would be interested in the mechanics of it, but throwing around jargonistic terms that most people on here wouldn't understand doesn't further the discussion. And while we don't know exactly what's going on, at least what is there is there and can be used to base rules around - rules that would mean we could make an unequivocal adjudication on a word, or at least much more frequently than is the case now. And even you think the results would be less "correct" than Susie's decisions, and made by some shadowy conspiratorial web developer, at least they'd be more consistent.

I can't let this slide. What is jargonistic about me mentioning the exact mechanism on which your entire system for validating words is based? At worst the casual reader who doesn't mind what your system is based on reads the word "redirect" and understands, and moves on. At best they can look it up and see what 302 exactly stands for, how it works, why it's such a bad idea to use it as for unequivocal adjudication, and why we would never know if they'd be consistent.

That's some weak shit from you.

[Jon O'Neill]

But, yes, we do need more from OUP. And what happened seems strange. Damian said they were clarifying stuff with them, but in the end nothing changed. I'm not sure what went wrong, but something did. I might send them another e-mail. Last time I did that, shortly afterwards they changed from the paper dictionaries to the ODO. And while I can't exactly claim credit, I would argue that there is a reasonable non-zero chance that it had some effect in terms of them getting their arses in gear. And if I do send another e-mail, as before I will post it in on the forum, so other people can copy it if they want to put the pressure on.

You say something must have gone wrong, but I imagine the conversation went something like:

DE "could we have a definitive list of what is or isn't a word, based on ODO headwords, as per the rules of Countdown, accounting for the following mistakes in your dictionary

• "
  
  OUP "Nope, sorry. Here's a list of what redirects to header entries. We got a few Eng. Lit. students in for a week to go through every headword and choose the valid inflections.. for new words we basically wing it and if anyone complains we just update it. Actually Mike from Finance ordered two plates of salmon the other night by saying 'two salmons please' so we added that lol! Ah, the finance team.. what a load of gad abouts! Anyway it later turned out that one of the temps was an imposter going under the name of Ecclesiastes Myanmar and decided to taint the list by redirecting random questionable plurals to weird pages on the website.. oh well, what are ya gonna do eh!"
  
  DE "ok, this is really going to upset Gevin, and it pretty much ruins our chances of getting Countdown into the Olympics, but we'll have to stick with the tried and trusted method until this situation changes."

[Gavin Chipper]

The reason I didn't have this debate with you before is that a lot of words are wasted on what is quite a simple problem. I like to argue with you so I entertained myself a little with talk of who said what and why, but it's completely irrelevant, and as I suspected, has given credence to your logically unsound thinking.

We are advocating two different systems. The drawbacks of the system I am advocating:

1. Mass noun/other rules are not applied properly
2. Mass noun/other rules are not applied consistently in individual cases, leading to uncertainty

Does your system fix either of these?

1. No. I have no idea who decided what is or isn't a valid inflection/comparative/superlative/plural form in the ODO, but it doesn't take long to find some very objectionable results. See ABOUTS; also SALMONS which is redirected despite the plural being specified as SALMON.
2. Who knows? Since the list of redirects is completely opaque, we wouldn't know if it did change on an individual basis. When ABOUTS and SALMONS get fixed, how do we know? There is no extra certainty.

So your system doesn't fix any of the current issues. So does your system have any drawbacks?

1. Opaque - things could be allowed or disallowed without us knowing why.

You've already agreed with this so I won't go on making other points.. they're unnecessary to prove that my system is better.

These things come in degrees. Is there any reason to suggest that the ODO is worse at applying the rules than Susie? Susie has allowed SALMONS, for example. On 1. I wouldn't necessarily say one is better than the other. ABOUTS is weird, but as I say, we can have additional rules that would disallow it.

But as for 2., yes, it's opaque, but we're told that there are four updates a year (I think), so that should apply to plurals of mass nouns etc. as much as whole entries. Actual errors might be changed as they're found, but these aren't likely to be that common. So yes, I would argue that using the ODO search box would lead to greater consistency and certainty.

But there are many ways of applying the rules, not simply what we have now and what I am suggesting. I made a suggestion , which is open to improvements, and you have argued against that and concluded that therefore what we have now is best. It doesn't follow.

[Gavin Chipper]

For a start, there might be some people who are interested in the mechanics behind the idea you are suggesting. I naively thought you might be one of those people. But never mind. What we know is that there is information somewhere which tells us what search terms redirect to what page. What we don't know is who decided what redirects where, and whether they are applying the guidance notes from the front of the ODE any better than Susie does. What we also don't know is whether they are even trying to apply those notes. What we also don't know is why some borderline entries redirect to a Legal Style Guide. What we also don't know is why some borderline entries redirect to completely different words.

Obviously I would prefer a more transparent system and would be interested in the mechanics of it, but throwing around jargonistic terms that most people on here wouldn't understand doesn't further the discussion. And while we don't know exactly what's going on, at least what is there is there and can be used to base rules around - rules that would mean we could make an unequivocal adjudication on a word, or at least much more frequently than is the case now. And even you think the results would be less "correct" than Susie's decisions, and made by some shadowy conspiratorial web developer, at least they'd be more consistent.

I can't let this slide. What is jargonistic about me mentioning the exact mechanism on which your entire system for validating words is based? At worst the casual reader who doesn't mind what your system is based on reads the word "redirect" and understands, and moves on. At best they can look it up and see what 302 exactly stands for, how it works, why it's such a bad idea to use it as for unequivocal adjudication, and why we would never know if they'd be consistent.

That's some weak shit from you.

Well no, you were chucking in the term as if it made it sound really bad as in "This system is shit, not because it redirects you [we knew that anyway], but because it uses the worst type of redirect - the HTTP 302!" It added nothing.

[Gavin Chipper]

But, yes, we do need more from OUP. And what happened seems strange. Damian said they were clarifying stuff with them, but in the end nothing changed. I'm not sure what went wrong, but something did. I might send them another e-mail. Last time I did that, shortly afterwards they changed from the paper dictionaries to the ODO. And while I can't exactly claim credit, I would argue that there is a reasonable non-zero chance that it had some effect in terms of them getting their arses in gear. And if I do send another e-mail, as before I will post it in on the forum, so other people can copy it if they want to put the pressure on.

You say something must have gone wrong, but I imagine the conversation went something like:

DE "could we have a definitive list of what is or isn't a word, based on ODO headwords, as per the rules of Countdown, accounting for the following mistakes in your dictionary

• "
  
  OUP "Nope, sorry. Here's a list of what redirects to header entries. We got a few Eng. Lit. students in for a week to go through every headword and choose the valid inflections.. for new words we basically wing it and if anyone complains we just update it. Actually Mike from Finance ordered two plates of salmon the other night by saying 'two salmons please' so we added that lol! Ah, the finance team.. what a load of gad abouts! Anyway it later turned out that one of the temps was an imposter going under the name of Ecclesiastes Myanmar and decided to taint the list by redirecting random questionable plurals to weird pages on the website.. oh well, what are ya gonna do eh!"
  
  DE "ok, this is really going to upset Gevin, and it pretty much ruins our chances of getting Countdown into the Olympics, but we'll have to stick with the tried and trusted method until this situation changes."

Maybe that is what happened.

[Jon O'Neill]

The reason I didn't have this debate with you before is that a lot of words are wasted on what is quite a simple problem. I like to argue with you so I entertained myself a little with talk of who said what and why, but it's completely irrelevant, and as I suspected, has given credence to your logically unsound thinking.

We are advocating two different systems. The drawbacks of the system I am advocating:

1. Mass noun/other rules are not applied properly
2. Mass noun/other rules are not applied consistently in individual cases, leading to uncertainty

Does your system fix either of these?

1. No. I have no idea who decided what is or isn't a valid inflection/comparative/superlative/plural form in the ODO, but it doesn't take long to find some very objectionable results. See ABOUTS; also SALMONS which is redirected despite the plural being specified as SALMON.
2. Who knows? Since the list of redirects is completely opaque, we wouldn't know if it did change on an individual basis. When ABOUTS and SALMONS get fixed, how do we know? There is no extra certainty.

So your system doesn't fix any of the current issues. So does your system have any drawbacks?

1. Opaque - things could be allowed or disallowed without us knowing why.

You've already agreed with this so I won't go on making other points.. they're unnecessary to prove that my system is better.

These things come in degrees. Is there any reason to suggest that the ODO is worse at applying the rules than Susie? Susie has allowed SALMONS, for example. On 1. I wouldn't necessarily say one is better than the other. ABOUTS is weird, but as I say, we can have additional rules that would disallow it.

No, it wasn't allowed. She offered it, but it's obvious (to me) that contestant declarations are subject to more scrutiny than DC offerings. I can't see SALMONS being allowed.

But as for 2., yes, it's opaque, but we're told that there are four updates a year (I think), so that should apply to plurals of mass nouns etc. as much as whole entries. Actual errors might be changed as they're found, but these aren't likely to be that common. So yes, I would argue that using the ODO search box would lead to greater consistency and certainty.

But there are many ways of applying the rules, not simply what we have now and what I am suggesting. I made a suggestion , which is open to improvements, and you have argued against that and concluded that therefore what we have now is best. It doesn't follow.

I never said it was perfect. I have concluded that what we have now is the best option that's been suggested so far. Until your originally concrete, now very vague, suggestion resolidifies, that will remain the case.

[Jon O'Neill]

Well no, you were chucking in the term as if it made it sound really bad as in "This system is shit, not because it redirects you [we knew that anyway], but because it uses the worst type of redirect - the HTTP 302!" It added nothing.

Just to reiterate my exact words, because I won't stand by and let you portray me as a cunt when I'm actually the voice of the people:

Your solution is poor because it relies on a database table of HTTP 302 redirects

The new information I'm adding is the underlying mechanism. It is important information because it distinguishes between something arbitrary stored somewhere which we can't get at, and some mechanism for determining whether each sense of the word can be (for example) pluralised based on what's actually in the definition, which actually WOULD be more consistent than the status quo! Yes, it's obvious to both of us that the former case is what's happening, but I've had to spell out so much quite obvious stuff to you that I thought it might be necessary.

[Jon Corby]

FYI, SALMONS was disallowed in the 30th BC I think, I was reading an old thread the other day and I'm sure I saw some discussion about it.

[Tracey Anne Mills]

FYI, SALMONS was disallowed in the 30th BC I think, I was reading an old thread the other day and I'm sure I saw some discussion about it.

That was in Chris Wills v Jack Hurst 30BC

[Gavin Chipper]

No, it wasn't allowed. She offered it, but it's obvious (to me) that contestant declarations are subject to more scrutiny than DC offerings. I can't see SALMONS being allowed.

Well, maybe it wasn't allowed as such but someone watching the show might later offer the word on the basis that Susie suggested it, so it's not consequence-free.

I never said it was perfect. I have concluded that what we have now is the best option that's been suggested so far. Until your originally concrete, now very vague, suggestion resolidifies, that will remain the case.

It's not very vague. It would mean we would disallow words that redirect to entries that they can't reasonably be a derivative of. And since we're only really talking about mass noun plurals and comparatives/superlatives, it would be quite simple to define. Noun + s or +es if it ends in an O etc. I can't be bothered to come up with a definitive set of rules, but it can easily be done. That's all you need to know.

[Jon O'Neill]

No, it wasn't allowed. She offered it, but it's obvious (to me) that contestant declarations are subject to more scrutiny than DC offerings. I can't see SALMONS being allowed.

Well, maybe it wasn't allowed as such but someone watching the show might later offer the word on the basis that Susie suggested it, so it's not consequence-free.

That could happen with any word validation system unless every offering is checked. I doubt she would've offered it if she'd had time to check it so the validation system is irrelevant to your point.

I never said it was perfect. I have concluded that what we have now is the best option that's been suggested so far. Until your originally concrete, now very vague, suggestion resolidifies, that will remain the case.

It's not very vague. It would mean we would disallow words that redirect to entries that they can't reasonably be a derivative of. And since we're only really talking about mass noun plurals and comparatives/superlatives, it would be quite simple to define. Noun + s or +es if it ends in an O etc. I can't be bothered to come up with a definitive set of rules, but it can easily be done. That's all you need to know.

As I asked earlier, what's your test of reasonability? Is it similar to the current test of reasonability for Mass Nouns? That doesn't sound like an improvement. To restate, there is already a definitive set of rules. So again your alternative is no better than the status quo.

[Gavin Chipper]

I never said it was perfect. I have concluded that what we have now is the best option that's been suggested so far. Until your originally concrete, now very vague, suggestion resolidifies, that will remain the case.

It's not very vague. It would mean we would disallow words that redirect to entries that they can't reasonably be a derivative of. And since we're only really talking about mass noun plurals and comparatives/superlatives, it would be quite simple to define. Noun + s or +es if it ends in an O etc. I can't be bothered to come up with a definitive set of rules, but it can easily be done. That's all you need to know.

As I asked earlier, what's your test of reasonability? Is it similar to the current test of reasonability for Mass Nouns? That doesn't sound like an improvement. To restate, there is already a definitive set of rules. So again your alternative is no better than the status quo.

As I said, it would be a specific list of common plurals and comparatives/superlatives. There's more than one possible way of doing it, and whether I bother to precisely define it makes no difference. I could do it, or someone else could, and that's what matters.

E.g. If they're found in the search but not specified, you could allow adding an S for a noun, also ES if it ends in O or I, or changing Y or EY to IES. And if it's an adjective, allow ER, EST to go on the end, or R, ST if it ends in E, a doubling of the final consonant, change the Y to an I before ER or EST etc.

Something like ABOUTS as a plural for GADABOUT would clearly be thrown out by this.

We could keep it simpler and say if it's headword + any combination of letters and it's found by the search box then allow it, and also allow the chopping off of certain letters from the end such as a Y before adding letters. In fact, unless you can find something glaringly stupid, I say go with this.

[Jon O'Neill]

SALMONS?

[Gavin Chipper]

SALMONS?

Fair enough. I don't think it would be completely insane to allow it, but if not, then we could still come up with other rules that are still objective.

We could use the search box for comparatives, superlatives and plurals of mass nouns, and have everything else covered by basic rules that involve just looking up the headword.

But anyway, I was quite happy for this discussion to happen when the ODO was first announced. It would have been worthwhile, and perhaps Countdown could have taken some of it aboard. I was never entrenched in my viewpoint, but you just kept disagreeing from the sidelines without making any points, until you suddenly decided to now.

Edit - The bottom line is that I still think the search box makes a better adjudicator of mass noun plurals than what we have.

[Jon O'Neill]

My points seem pretty obvious. We agree on where we disagree now, which is fine. You think the search box, i.e. an unknown adjudicator, is better than what we have now. I disagree.

As an aside, another confusion is accented words. If you type "abbe" in the search box you get redirected somewhere mad. If you type "abbé", you get where you need to go. If you type "abbes" you get where you need to go. "Deracine" takes you to déraciné, so that works. Why doesn't "abbe"?

In Countdown you don't need to declare whether or not your word has an accent so it seems crazy to disallow a word on that basis.

As you can see, I'm having a look through alphabetically and it's not taken me long to find new, weird stuff, which is exactly what I expected from your hidden adjudicator.

[Matt Morrison]

Can't think about Gevin's hidden adjudicator and keep a straight face.

[Mark Deeks]

Remember Co:Lin 2014? His adjudicator wasn't hidden that night. Loud and proud.

[Matt Morrison]

[Gavin Chipper]

Nice phone cover. I might get one for my phone, but I'll have to check it's big enough first.

By the way, Jono might be giving it all this in this thread, but let's not forget that he was wandering around with shit in his hair the other day and needed me to save him.

[Adam Dexter]

SALMONS?

Surely SALMONS is allowed in the restaurant rule?

[Graeme Cole]

SALMONS?

Surely SALMONS is allowed in the restaurant rule?

No. The food rule just means SALMON can be pluralised. The plural is still SALMON.

[Gavin Chipper]

SALMONS?

Surely SALMONS is allowed in the restaurant rule?

No. The food rule just means SALMON can be pluralised. The plural is still SALMON.

Exactly. I mean, someone might ask for two "salmons" in a restaurant, but people say lots of things that aren't in the ODO. The ODO has the plural as SALMON and that applies to all uses of the plural.

Anyway, Jon O'Neill has run a program checking a much larger selection of words on the ODO (he asked for my help in how best to do that though) to see what the search box comes up with. It turns out that it is pretty poor, and he rang me up to brag about this and wanted me to post on here backing down and issuing a full apology or something (as if that would ever happen).

For example, it doesn't find ARCHAEAL, even though it's explicitly listed under ARCHAEA. It doesn't find WOODBINES, even though WOODBINE is just a normal noun. It also redirects to some nonsense when you type in ASSIGNOR, even though it's listed under ASSIGN. And it can't find AM.

But anyway, I'm quite happy to accept that the search box has been found wanting. I always said from the start that the search box would be a good method of adjudication, but only if it didn't come up with rubbish results. So to me, this is an empirical finding. But this is where we still disagree. Jono thinks it was obvious from the start and didn't even need testing, but he just decided to do so anyway to prove me wrong. But we didn't know who was responsible for the redirects. It could have been a properly thought out system. It just turned out not to be.

Plus, you could still use it just on the plurals of mass nouns to make it consistent...

[Jon O'Neill]

No that would be ludicrous. If you can't trust it to adjudicate plurals of count nouns then how can you trust it not to come up with freaky adjudications for mass nouns?

[Clive Brooker]

It could have been a properly thought out system. It just turned out not to be.

Even if it were, it wouldn't automatically follow that it always means what you think it ought to. For example, CLEAROUT redirects to clear-out, thereby showing you the error of your ways rather than condoning the use of the unhyphenated version.

[Gavin Chipper]

It could have been a properly thought out system. It just turned out not to be.

Even if it were, it wouldn't automatically follow that it always means what you think it ought to. For example, CLEAROUT redirects to clear-out, thereby showing you the error of your ways rather than condoning the use of the unhyphenated version.

Yep. I mentioned this before in another thread. There's an entry for STRUCTURAL STEEL and it will find it if you search for STRUCTURALSTEEL. So the idea was always to see how it works and fits the rules around it accordingly.

[Tracey Anne Mills]

The challenger was brilliant TV, her conundrum buzz properly cracked me up

I enjoyed her comments during the selections.

Agreed. Mickie was amazing!

She was a nice person