Co-Event Suggestion Box

[Conor]

Does anybody have some low effort high impact suggestions for reducing the % of botched printed conundrum reveals?

I had a botched conundrum given to me in at least 5/8 of the events I attended in 2023 which had paper conundrums.

Reasons for botched conundrums reveals could be any of (but not limited to):
- Wonky
- Upside down (correctly oriented relative to host rather than players)
- Face down
- Accidentally revealed while setting up to reveal
- Stuck to a clammy hand when the hand is pulled away
- Blown over by a gust of air caused by the action of revealing
- Answer still attached!

I accept that botched conundrums will always happen, but I am sure there's something simple we can do to reduce it significantly!

Another possible way to botch: Not having the timer ready when you reveal.

Trying to juggle the conundrum reveal with starting the timer can be tricky. I'd rather the host focused on getting a clean reveal than botching it in order to start the timer right away. Most conundrums are solved either very quickly, with a guess that's pegged just inside the timer end, or not at all. Whether it's 28 or 32 seconds thinking time makes little difference really.

[Gavin Chipper]

no crisps before conundrums

It's OK if you've got a spoon.

[JackHurst]

no crisps before conundrums

It's OK if you've got a spoon.

The Mellor method

[JackHurst]

Just looking at the sign ups for Colin. Rob and Jack W very likely to be Table 1 together for round 2. Therefore it's a sound strategy to try and win badly in R1 to get on table 2 or 3 for R2. Would be cool at tourneys like this is r2 was random game vs somebody on same wins.

[Fiona T]

Just looking at the sign ups for Colin. Rob and Jack W very likely to be Table 1 together for round 2. Therefore it's a sound strategy to try and win badly in R1 to get on table 2 or 3 for R2. Would be cool at tourneys like this is r2 was random game vs somebody on same wins.

I think deliberately throwing points would be pretty poor sportsmanship, and could definitely affect a player's chances of making the final if people are on the same wins after their games. (I also don't think any of the top players would actually do this, although if they're well ahead they might risk the dodgy 8/9, but then most of us would!)

But I appreciate that wasn't your main point Surely if those players don't meet in round two, then they're likely to meet in round 3?

[JackHurst]

But I appreciate that wasn't your main point Surely if those players don't meet in round two, then they're likely to meet in round 3?

If you are #4 after R1 then your R3 table can only have one of the top 3 players. When it's players like Rob and Jack W the difference between having to play both of them vs just one of them is pretty big.

[Gavin Chipper]

If it were up to me, we'd bin off the Swiss pairs/triples system entirely. I've discussed this general thing before with various tweaks, but here is my definitive system that I think would work:

The first round is random. A lot of events do this anyway, although Ben traditionally likes to do his annual trolling seed pot system at COLIN.

For later rounds, instead of trying to put people against players who have done similarly, you instead give people opponents that mean they have faced the range of opposition available to the best extent possible, with the exception that unbeaten players still face each other, so we don't end up with a load of unbeaten players at the end.

So how does this work in practice? Let's say there are 20 people at an event, with three players to a table. John plays two players at random in round one. After round one, John is ranked 10th, with his opponents 5th and 18th.

For every player, an Average Opponent Rank is calculated. The player concerned is not included in the ranking list for these purposes. So on John's opponent list, he had played 5 and 17 (not 18). This gives an average of 12 for John. Each player's opponent list has 19 players on, so the overall average opponent rank is 10. For the next round, the players will be sorted into triples that minimise the standard deviation of each player's Average Opponent Rank, with the following condition:

Unbeaten players face each other. If there are a multiple of 3 unbeaten players, put them on tables of 3 together. If it's a multiple of 3 minus 1, you'd have 3s and a single 2. If it's multiple of 3 plus 1, I think it would make sense to have two tables of 2 and the rest 3s, rather than having a single unbeaten player avoiding all the others.

The unbeaten players aren't completely separate from the Average Opponent Rank calculation. The calculation is still done across all players and the standard deviation is minimised but with the unbeaten player condition.

Ranks change across the tournament, and Average Opponent Rank always uses the current rank of each player. So when John's round 3 table is determined, his opponents in round 1 won't be stuck on 5 and 17, but what they are on after round 2.

Any questions?

Edit - That's for triples. For tables of two (e.g. Bristol style), it would be the same basic system, but just simpler because everyone is given a single opponent to get them as near as possible to the mean Average Opponent Rank.

Edit 2 - As with the Swiss system, if this is unfeasible computationally, you just use some "greedy" algorithm that gets close enough.

[Fiona T]

If it were up to me, we'd bin off the Swiss pairs/triples system entirely. I've discussed this general thing before with various tweaks, but here is my definitive system that I think would work:

The first round is random. A lot of events do this anyway, although Ben traditionally likes to do his annual trolling seed pot system at COLIN.

For later rounds, instead of trying to put people against players who have done similarly, you instead give people opponents that mean they have faced the range of opposition available to the best extent possible, with the exception that unbeaten players still face each other, so we don't end up with a load of unbeaten players at the end.

So how does this work in practice? Let's say there are 20 people at an event, with three players to a table. John plays two players at random in round one. After round one, John is ranked 10th, with his opponents 5th and 18th.

For every player, an Average Opponent Rank is calculated. The player concerned is not included in the ranking list for these purposes. So on John's opponent list, he had played 5 and 17 (not 18). This gives an average of 12 for John. Each player's opponent list has 19 players on, so the overall average opponent rank is 10. For the next round, the players will be sorted into triples that minimise the standard deviation of each player's Average Opponent Rank, with the following condition:

Unbeaten players face each other. If there are a multiple of 3 unbeaten players, put them on tables of 3 together. If it's a multiple of 3 plus 1, you'd have 3s and a single 2. If it's multiple of 3 minus 1, I think it would make sense to have two tables of 2 and the rest 3s, rather than having a single unbeaten player avoiding all the others.

The unbeaten players aren't completely separate from the Average Opponent Rank calculation. The calculation is still done across all players and the standard deviation is minimised but with the unbeaten player condition.

Ranks change across the tournament, and Average Opponent Rank always uses the current rank of each player. So when John's round 3 table is determined, his opponents in round 1 won't be stuck on 5 and 17, but what they are on after round 2.

Any questions?

Edit - That's for triples. For tables of two (e.g. Bristol style), it would be the same basic system, but just simpler because everyone is given a single opponent to get them as near as possible to the mean Average Opponent Rank.

Edit 2 - As with the Swiss system, if this is unfeasible computationally, you just use some "greedy" algorithm that gets close enough.

In theory the basic system sounds 'fairer', but not for players at the top. If 100% fairness is the aim, the best players shouldn't be penalised for being the best. I imagine it would be pretty disheartening for lower ranked players, who, instead of playing other low ranked players have to play the range. For players like me and you who are typically in the middle of the pack, it's probably business as usual. Given the swiss system already pairs unbeaten players, and lower ranked players have a lot more fun playing each other and rarely trouble the top players the current system appears to work. I can think of one or two tournaments where a surprise player has made the final, but that's all for the good I think!

Edit to add - I always stick on my events "The primary aim is to have fun!" - I think this is best served by mainly playing opponents with a similar ability to oneself!

[Marc Meakin]

I feel there is less risk for the elite players in Co Events
The only real element of luck is that standard letter formations and easy number targets leave it as a conundrum shoot out
where as in Scrabble an elite player can lose to practicly anyone if the tile fairy is especially unkind.

Maybe having round Robin games between Elie players would eliminate any risks in both Co events and scrabble

[Gavin Chipper]

In theory the basic system sounds 'fairer', but not for players at the top. If 100% fairness is the aim, the best players shouldn't be penalised for being the best.

It's supposed to be a compromise system. As fair as possible, while not having too many unbeaten players at the end of the day. In fact, that is also a fairness thing. It can be seen as unfair that an unbeaten player doesn't make the final. But anyway, that's the point of it.

I imagine it would be pretty disheartening for lower ranked players, who, instead of playing other low ranked players have to play the range. For players like me and you who are typically in the middle of the pack, it's probably business as usual. Given the swiss system already pairs unbeaten players, and lower ranked players have a lot more fun playing each other and rarely trouble the top players the current system appears to work. I can think of one or two tournaments where a surprise player has made the final, but that's all for the good I think!

Edit to add - I always stick on my events "The primary aim is to have fun!" - I think this is best served by mainly playing opponents with a similar ability to oneself!

I don't buy the whole thing that people only want to play people at a similar level to themselves. I certainly don't. Plus the Swiss system still throws out annoyances anyway. For example, if someone loses their first two games but they play well, they might think that at least they'll get easier games next time. Well, it often doesn't work out that way because they can end up on a table with other players in the same position. Someone else on a win and a loss might get an easier draw.

Also, this post from me sums up my feelings on the matter quite well. So I'll quote it wholesale. The part about Prune is important. Playing Prune isn't fun. The primary aim is to have fun. Prune should be shared round more evenly. And I'm pleased with my use of the word "foist" there.

When people criticise this and that, one thing that often seems to escape criticism (except from me) is the Swiss pairs system itself. So I'll summarise:

1. It's intrinsically biased, favouring those who have done poorly so far. This is a mathematical fact.

2. It patronisingly assumes that people only want to play people of a similar level to them and can't possibly play one of the top players unless they are one themselves (except that in events like COLIN the first round is run to different logic, questioning the commitment to that philosophy anyway).

3. It means that weaker players are more likely to have the ignominy of playing Prune. It's annoying having to play Prune even if it's a free win. People don't want to play Prune and it's unfair to just foist it on the players who've done badly (which often ends up being the same people at many events) rather than spread it around.

And from Mark Deeks in this thread:

ADDITIONAL - At said Co:Lins, my favourite round is always the first one, where I'll play players I wouldn't normally.

[Fiona T]

I don't disagree that players who have done a bit of apto practice can't have an enjoyable game against a top player - my first table at my first event was against Callum and Robbo - both beat me thoroughly, but I scored on a few rounds and had a good time. I've had some very enjoyable games against Jack, Bradley, Rob, Callum, Maus etc. But if you get a complete newbie who gets completely thrashed that's a different case. There have been players who have scored 0 at events. If that happens to you once it's pretty horrible - if it happens 3 or 4 times, you're almost certainly not coming back. It's pretty horrible for the person doing the beating too. I do agree about prune - I think ideally event organisers would rope in a friend or family member to make the numbers round if needed, but easier said than done (my family certainly are not obliging!)

I think the current system works though - I've been at tournaments where I've been on the bottom table on round 3 and tournaments where I've been near the top. I never feel like I'm only playing players the same level as me, but where I've had a terrible start I get an easier draw, and where I'm smashing it it's tougher - that seems right to me!

[Adam S Latchford]

Just looking at the sign ups for Colin. Rob and Jack W very likely to be Table 1 together for round 2. Therefore it's a sound strategy to try and win badly in R1 to get on table 2 or 3 for R2. Would be cool at tourneys like this is r2 was random game vs somebody on same wins.

i've legitimately tried this as a strategy at birmingham last year

was second or third seed after the first 6 games but the strategy caught up with me and lost my final 2 games to come like 11th

Also with it being 9 rounders avoiding rob/jack/jack doesn't mean your getting an easy game

[JackHurst]

Just looking at the sign ups for Colin. Rob and Jack W very likely to be Table 1 together for round 2. Therefore it's a sound strategy to try and win badly in R1 to get on table 2 or 3 for R2. Would be cool at tourneys like this is r2 was random game vs somebody on same wins.

i've legitimately tried this as a strategy at birmingham last year

was second or third seed after the first 6 games but the strategy caught up with me and lost my final 2 games to come like 11th

Also with it being 9 rounders avoiding rob/jack/jack doesn't mean your getting an easy game

Why did you try this at an 8 game event

The extra round balances things out anyway so you are just costing yourself points haha

[Adam S Latchford]

Just looking at the sign ups for Colin. Rob and Jack W very likely to be Table 1 together for round 2. Therefore it's a sound strategy to try and win badly in R1 to get on table 2 or 3 for R2. Would be cool at tourneys like this is r2 was random game vs somebody on same wins.

i've legitimately tried this as a strategy at birmingham last year

was second or third seed after the first 6 games but the strategy caught up with me and lost my final 2 games to come like 11th

Also with it being 9 rounders avoiding rob/jack/jack doesn't mean your getting an easy game

Why did you try this at an 8 game event

The extra round balances things out anyway so you are just costing yourself points haha

I agree, misstep in an 8

you live and learn

[Matt Rutherford]

The Pot Draw Method
This has come from running a chess club in a school, but how it runs at a Co-event would be interesting. Not saying it's perfect, but would be interested to see people's ideas on it. It's similar(ish) to some of what has been mentioned before. For context, there are three games at each club meeting.

-First game is entirely random.
-Second game, you split results. Winners into winners pot, losers into losers pot. You draw the winners to play each other and losers to play each other.
-Third game, do the same for the results of the second game.

So for Co-Events, you basically rinse-and-repeat for Bristol. For Lincoln...
-First round, draw at random
-Second round, separate into three groups-those who won two on first table, those who one, those who won none.
-Third round-repeat that looking only at how they did in the Second Round. So those who won 2 games in the Second Round go into one pot. Those who won 1 game in the Second Round into another and so on...Repeat for further rounds, basing the draws only how they did in the previous rounds.

(N.B-You draw from the pots as much as possible, but when there aren't enough to fill a three-way table in one pot, you draw from the next lowest pot to fill it. This may need refining if it was ever to be used)

Thoughts?

More generally, no system will be perfectly fair, or allow for perfect variation, and any system can be gamed. The one I've outlined above isn't perfect, and I'm sure has holes. HOWEVER, whatever the system, as long as it doesn't utilize either...
-Entirely random draws at every point
-Divisions at any point

Then it's a good system. If you avoid those two methods like an irritating relative with the plague, then job's a good-un

[JackHurst]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

[JackHurst]

If it were up to me, we'd bin off the Swiss pairs/triples system entirely. I've discussed this general thing before with various tweaks, but here is my definitive system that I think would work:
...

The Pot Draw Method
This has come from running a chess club in a school, but how it runs at a Co-event would be interesting. Not saying it's perfect, but would be interested to see people's ideas on it. It's similar(ish) to some of what has been mentioned before. For context, there are three games at each club meeting.
....

Both of these methods would be interesting, and probably an improvement on Swiss in terms of variety and fairness. Perhaps the downsides are:
- Difficult to explain clearly in less than 30s at the start of the event
- Porting the logic across to tools like Atropine might be fiddly.


Perhaps we've been too focused on implementation and it would be more productive to agree some key principles first.

If I had to pick 5 principles I would go for:
- Avoid rematches
- Minimise the number of players on 0% or 100%
- The final round of games should most likely pit the best players against each other.
- Over the course of a season, if I go to events regularly I should face a wide variety of players.
- Hosts should be able to explain the system to attendees in 1 minute or less. (I.e it's not overly complicated)

Edit: I thought of a 6th principle, so now I need to whittle my 6 down to 5. 6th one was order of results should not matter, so a LLLWWW or WWWLLL should be the same.

[Ben Wilson]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

They used to. Adam 'awesome' Beach has gone through the packs and standardised them all, so everyone from COLIN onward should be playing to the same set regardless of which table they're on.

[Fiona T]

If it were up to me, we'd bin off the Swiss pairs/triples system entirely. I've discussed this general thing before with various tweaks, but here is my definitive system that I think would work:
...

The Pot Draw Method
This has come from running a chess club in a school, but how it runs at a Co-event would be interesting. Not saying it's perfect, but would be interested to see people's ideas on it. It's similar(ish) to some of what has been mentioned before. For context, there are three games at each club meeting.
....

Both of these methods would be interesting, and probably an improvement on Swiss in terms of variety and fairness. Perhaps the downsides are:
- Difficult to explain clearly in less than 30s at the start of the event
- Porting the logic across to tools like Atropine might be fiddly.


Perhaps we've been too focused on implementation and it would be more productive to agree some key principles first.

If I had to pick 5 principles I would go for:
- Avoid rematches
- Minimise the number of players on 0% or 100%
- The final round of games should most likely pit the best players against each other.
- Over the course of a season, if I go to events regularly I should face a wide variety of players.
- Hosts should be able to explain the system to attendees in 1 minute or less. (I.e it's not overly complicated)

Edit: I thought of a 6th principle, so now I need to whittle my 6 down to 5. 6th one was order of results should not matter, so a LLLWWW or WWWLLL should be the same.

Which, if any, of those 5/6 principles does the current system fail on?

(I also like the sound of 'pot draw', but not yet convinced that the current system needs reform!)

[JackHurst]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

They used to. Adam 'awesome' Beach has gone through the packs and standardised them all, so everyone from COLIN onward should be playing to the same set regardless of which table they're on.

Good to hear it. Now we just need to make sure that attendees look after them properly

[JackHurst]

If it were up to me, we'd bin off the Swiss pairs/triples system entirely. I've discussed this general thing before with various tweaks, but here is my definitive system that I think would work:
...

The Pot Draw Method
This has come from running a chess club in a school, but how it runs at a Co-event would be interesting. Not saying it's perfect, but would be interested to see people's ideas on it. It's similar(ish) to some of what has been mentioned before. For context, there are three games at each club meeting.
....

Both of these methods would be interesting, and probably an improvement on Swiss in terms of variety and fairness. Perhaps the downsides are:
- Difficult to explain clearly in less than 30s at the start of the event
- Porting the logic across to tools like Atropine might be fiddly.


Perhaps we've been too focused on implementation and it would be more productive to agree some key principles first.

If I had to pick 5 principles I would go for:
- Avoid rematches
- Minimise the number of players on 0% or 100%
- The final round of games should most likely pit the best players against each other.
- Over the course of a season, if I go to events regularly I should face a wide variety of players.
- Hosts should be able to explain the system to attendees in 1 minute or less. (I.e it's not overly complicated)

Edit: I thought of a 6th principle, so now I need to whittle my 6 down to 5. 6th one was order of results should not matter, so a LLLWWW or WWWLLL should be the same.

Which, if any, of those 5/6 principles does the current system fail on?

(I also like the sound of 'pot draw', but not yet convinced that the current system needs reform!)

Different event formats probably fall short on different combinations of those principles. We're talking about the Colin format here then I think variety of opponent is the main drawback for me. I can't speak for everybody, but my own experience over the past couple of seasons has been very repetitive.

Are there any principles you would add/remove?

One thing I am getting from this thread is that there is an appetite to try the "random opponent on same number of wins" method at some event soon. I would hold off for a 6 game event though. 8 games has the benefit of the final round giving some variation and a bit of restoration to the balance anyway. 4 is always a bit of a shoot out.

[Fiona T]

Different event formats probably fall short on different combinations of those principles. We're talking about the Colin format here then I think variety of opponent is the main drawback for me. I can't speak for everybody, but my own experience over the past couple of seasons has been very repetitive.

Interesting - I guess players who generally win (or lose) most of their games have a different experience to those of us who are a bit more unpredictable! Looking at the last 3 events I attended, of my 20 opponents, there was only one I played twice!

[Gavin Chipper]

- The final round of games should most likely pit the best players against each other.

Which, if any, of those 5/6 principles does the current system fail on?

The Swiss system used at COLIN and other players doesn't really do this. As far as I understand the top players after round 1 go against each other in round 2. So if these really are the top players, then in round 3, they'll be against different players.

Is this right - at COLIN, would ranks 1, 2 and 3 go against each other in the second round (assuming it didn't cause any problems such as repeated games elsewhere)?

[JackHurst]

Different event formats probably fall short on different combinations of those principles. We're talking about the Colin format here then I think variety of opponent is the main drawback for me. I can't speak for everybody, but my own experience over the past couple of seasons has been very repetitive.

Interesting - I guess players who generally win (or lose) most of their games have a different experience to those of us who are a bit more unpredictable! Looking at the last 3 events I attended, of my 20 opponents, there was only one I played twice!

I just went through my co event notebook and totalled all of my 2023 games

Including GFs and Focal EOY Finals
68 Games played across 10 events. 37 unique opponents
Maus 6
TCap 5
George 4
Chris M 3
Rob 3
JDS 3
Dan B 3
Darren G 2
MSR 2
Tim D 2
Adam L 2
Jack W 2
Stephen R 2
Ahmed 2
Tim H 2


Exlcluding GFs and Focal EOY Finals
55 Games played across 9 events. 37 unique opponents
Maus 4
TCap 4
JDS 3
Chris M 3
George 2
Rob 2
Dan B 2
Darren G 2
MSR 2
Stephen R 2

I don't know whether to think thats a lot of rematches, or not very many. What I do know is that I like TCap and Maus, but I am very bored of playing people called Tom.

[Thomas Cappleman]

- The final round of games should most likely pit the best players against each other.

Which, if any, of those 5/6 principles does the current system fail on?

The Swiss system used at COLIN and other players doesn't really do this. As far as I understand the top players after round 1 go against each other in round 2. So if these really are the top players, then in round 3, they'll be against different players.

Is this right - at COLIN, would ranks 1, 2 and 3 go against each other in the second round (assuming it didn't cause any problems such as repeated games elsewhere)?

That's right - and I think if at least 3 people are on 2 wins and it's a normal sized field it's incredibly unlikely that Atropine would choose to change that as its way of preventing repeats.

It's unlikely though that the top 3 after 1 round of Lincoln style will be the actual strongest 3 (given different random rounds to play), so most likely 2 of the very top players will still be on the top table for the following round.

[Fiona T]

Different event formats probably fall short on different combinations of those principles. We're talking about the Colin format here then I think variety of opponent is the main drawback for me. I can't speak for everybody, but my own experience over the past couple of seasons has been very repetitive.

Interesting - I guess players who generally win (or lose) most of their games have a different experience to those of us who are a bit more unpredictable! Looking at the last 3 events I attended, of my 20 opponents, there was only one I played twice!

I just went through my co event notebook and totalled all of my 2023 games

Including GFs and Focal EOY Finals
68 Games played across 10 events. 37 unique opponents
Maus 6
TCap 5
George 4
Chris M 3
Rob 3
JDS 3
Dan B 3
Darren G 2
MSR 2
Tim D 2
Adam L 2
Jack W 2
Stephen R 2
Ahmed 2
Tim H 2


Exlcluding GFs and Focal EOY Finals
55 Games played across 9 events. 37 unique opponents
Maus 4
TCap 4
JDS 3
Chris M 3
George 2
Rob 2
Dan B 2
Darren G 2
MSR 2
Stephen R 2

I don't know whether to think thats a lot of rematches, or not very many. What I do know is that I like TCap and Maus, but I am very bored of playing people called Tom.

The finals do skew things - at an event with a final you're almost certainly gonna be playing someone you've played, and the focal finals by their very nature will mean probably all are rematches, 37 different opponents doesn't sound too bad, and the 4 rematches probably feels like a lot more because of GFs/Finals, but yeah 4 is still a lot.

[JackHurst]

It's unlikely though that the top 3 after 1 round of Lincoln style will be the actual strongest 3 (given different random rounds to play), so most likely 2 of the very top players will still be on the top table for the following round.

I'm inclined to disagree with this. Any single combination isn't very likely because there are lots of combinations possible, however, the table of the three strongest players is the most likely table among all tables. Under most systems, this would still be the case, because those three players are the three most likely to be on 2/2

[Thomas Cappleman]

It's unlikely though that the top 3 after 1 round of Lincoln style will be the actual strongest 3 (given different random rounds to play), so most likely 2 of the very top players will still be on the top table for the following round.

I'm inclined to disagree with this. Any single combination isn't very likely because there are lots of combinations possible, however, the table of the three strongest players is the most likely table among all tables. Under most systems, this would still be the case, because those three players are the three most likely to be on 2/2

It's the most likely, yes, but that's far from making it likely. There's a high chance that one of the next few strongest players will get a nine or two and so be in the top 3 after one round. Gevin's original post was about the top 3 being all together for round 2, and so all separate for round 3. That's incredibly unlikely to happen - and looking through a selection of recent events with full round results confirms that.

[JackHurst]

That's incredibly unlikely to happen - and looking through a selection of recent events with full round results confirms that.

Tom, I'm not sure where you are getting your data from to make this claim of "incredibly unlikely to happen" from, but I think you are just wrong here. It literally happened at the first 2 6 game events of the focal Calendar last season. I remember because I was on the tables!

Colin T1 R2 Me and Conor and Tim Down
Coman T1 R2 Me and Rob and Jack Worsely

[Thomas Cappleman]

That's incredibly unlikely to happen - and looking through a selection of recent events with full round results confirms that.

Tom, I'm not sure where you are getting your data from to make this claim of "incredibly unlikely to happen" from, but I think you are just wrong here. It literally happened at the first 2 6 game events of the focal Calendar last season. I remember because I was on the tables!

Colin T1 R2 Me and Conor and Tim Down
Coman T1 R2 Me and Rob and Jack Worsely

Are we arguing about the same thing? I'm saying that it's unlikely that the 3 strongest players in the room will be the actual top 3 after round 1 (and so on the same table for round 2, as that is basically guaranteed). And with all respect to Tim and his Hangover win last year, you wouldn't put him in the top 3 strongest players at Colin last year.

I will give you Coman though, that wasn't one of the ones I looked at (as I wasn't there so didn't know the format). That table is legit. I looked at Colins and Colons, where 2022 Colon had Conor and Ahmed on table 3 in round 2.

[JackHurst]

M8 I'm telling you I dread R2 fixtures more than R3 fixtures.

[Adam S Latchford]

where 2022 Colon had Conor and Ahmed on table 3 in round 2.

and some drunk absolute mug who actually bottled both games

[Gavin Chipper]

The question really is whether putting 1, 2, 3 against each other in round 2 is the best way to get the best players against each other in the final round. It doesn't seem likely.

It's not putting a value judgement on it. I'm just addressing Jack's list.

[Ben Wilson]

Thought I'd weigh in here with a bit of statistical data (and fwiw, I'm in favour of the 'draw 3 randomly on the same number of wins for round 2' idea) so here are the players on table 1 after rounds 2 and 3 of COLIN from 2005-2023, along with their eventual finishing positions.

Code: Select all

Round 2
2005* Chris Wills 1 Stewart Holden 8 Kevin Thurlow 18
2006  Stewart Holden 2 Kirk Bevins 5 Mike Brown 22
2007  Chris Wills 1 Paul Howe 2 Ben Wilson 7
2008  Chris Wills 3 Mark Tournoff 4 Stewart Holden 6
2009  Jon O'Neill 2 Dan Vanniasingham 3 Nick Wainwright 8
2010  Paul Howe 2 Jack Hurst 3 Richard Priest 11
2011  James Robinson 2 Mark Tournoff 9 Matt Bayfield 10
2012  Dinos Sfyris 1 Mark Deeks 5 Graeme Cole 24
2013  Jon O'Neill 7 Jack Worsley 10 James Robinson 20
2014  Jack Hurst 1 Kirk Bevins 3 Jon Corby 18
2015  Innis Carson 1 Giles Hutchings 3 Dylan Taylor 6
2016  Jack Worsley 1 Conor Travers 7 Zarte Siempre 17
2017  Dan McColm 6 Jack Worsley 7 Jen Steadman 10
2018  Jack Worsley 1 Callum Todd 2 Graeme Cole 17
2019  Jack Worsley 2 Rob Foster 3 Jon O'Neill 8
2020  Rob Foster 5 Jean Steadman 20 Jonathan de Souza 21
2022  Bradley Horrocks 1 Conor Travers 2 Dinos Sfyris 7
2023  Conor Travers 2 Jack Hurst 8 Tim Down 9

Code: Select all

Round 3
2005* Chris Wills 1 Mark Tournoff 2 Jim Bentley 7
2006  Mark Tournoff 3 Conor Travers 4 Kirk Bevins 5
2007  Paul Howe 2 Kirk Bevins 5 Michael Bowden 8
2008  Conor Travers 1 Chris Wills 3 Dan Vanniasingham 5
2009  Jon Corby 1 Nick Wainwright 8 Charlie Reams 13
2010  Paul Howe 2 Martin May 5 David O'Donnell 6
2011  Kirk Bevins 1 James Robinson 2 Josh Hurst 12
2012  Dinos Sfyris 1 Jon Corby 2 James Robinson 7
2013  Mark Deeks 1 Ed McCullagh 3 Jack Worsley 10
2014  Jack Hurst 1 Dylan Taylor 4 Rob Foster 13
2015  Mark Deeks 2 Jack Worsley 4 Graeme Cole 15
2016  Jack Worsley 1 Rob Foster 4 Andy Platt 14
2017  Dylan Taylor 1 Zarte Siempre 5 Dan McColm 6
2018  Jack Worsley 1 Thomas Cappleman 4 Ryan Taylor 12
2019  Bradley Horrocks 1 Callum Todd 6 Mark Deeks 11
2020  Jonathan Wynn 1 Rob Foster 5 Tom Carey 11
2022  Bradley Horrocks 1 Oliver Garner 3 Matthew Brockwell 9
2023  Callum Todd 1 Thomas Cappleman 5 Jack Hurst 8

*2005 was drawn differently to other years.
From 2012 onwards, finishing positions doesn't include grand final match.

Not sure what, if anything, this proves.

[Matt Rutherford]

Different event formats probably fall short on different combinations of those principles. We're talking about the Colin format here then I think variety of opponent is the main drawback for me. I can't speak for everybody, but my own experience over the past couple of seasons has been very repetitive.

Are there any principles you would add/remove?

One thing I am getting from this thread is that there is an appetite to try the "random opponent on same number of wins" method at some event soon. I would hold off for a 6 game event though. 8 games has the benefit of the final round giving some variation and a bit of restoration to the balance anyway. 4 is always a bit of a shoot out.

There's an appetite to try something, so the next thing to do is...Try. Something. This season seems good, because...

-Lincoln, Liverpool, Oxford are six-rounds, as normal. (Our 'control' tournaments in this case)
-Rugby, Sheffield, Durham are going for eight-rounds (Eight seems to even things out, but it needs some more testing)
-Reading doing more rounds owing to it being Bristol classroom-style (See above and apply)

So there'll be a comparison into how much six and eight games even things out.

Goat/Co Brum will be back in August (will open sign-ups post-Rugby), so some newer stuff could be trialled. From what's written above, my initial thoughts would be...

-Have both Goat and Co at six games.
-For one, use Draw Against Someone On Same Number Of Wins (the 'Geneva' method).
-For another, use the Pot Draw Method (the 'Zurich'* Method).

We'll have at least tried. There may be one formula people prefer. Or maybe the variation would make things more interesting. We don't know till we try. Offer is there and feedback is welcome!

(*These aren't compulsory names, but at least something will be slightly Swiss. Others may be better)

[Callum Todd]

Goat/Co Brum will be back in August (will open sign-ups post-Rugby), so some newer stuff could be trialled. From what's written above, my initial thoughts would be...

-Have both Goat and Co at six games.

Wouldn't a 6 game event on a Friday night be a bit long?

[JackHurst]

Is you had 4 players on 4 wins after 2 rounds of games how would you want the tables to be for rd3?

[Matt Rutherford]

Is you had 4 players on 4 wins after 2 rounds of games how would you want the tables to be for rd3?

Zurich pots-Separate them out depending on who won 2, who won 1, who won none in Rd 2. Draw from the pots.

Geneva wins-Three of the four on 4 wins play each other. The other 4 winner plays two who have won 3, selected all at random is how I would envision it

[Fiona T]

Is you had 4 players on 4 wins after 2 rounds of games how would you want the tables to be for rd3?

Zurich pots-Separate them out depending on who won 2, who won 1, who won none in Rd 2. Draw from the pots.

They've all won 2, but I guess there could be other players on 3 and 2 wins (who'd won both their R2 games) who would be in the same pot. Sounds like it's worth a try

[JackHurst]

I guess the question wasn't about method but more about outcome. What's better for R3 tables
Version A
T1 4W 4W 4W
T2 4W 3W 3W

Or
Version B
T1 4W 4W 3W
T2 4W 4W 3W

Version B has a big downside that 2 players on 3 wins are getting lumped with a pair of 4W players. Version A has the downside that one of the unbeaten players gets away with playing none of the others. Both feel quite unsatisfactory!

[Gavin Chipper]

I've thought about that before, and I think it's probably better overall to have 2 tables of 2 unbeaten players, rather than a 3 and a 1.

[Ben Wilson]

The upside of version B is that the 3W players will be the two with the highest aggregate score.

[JackHurst]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

They used to. Adam 'awesome' Beach has gone through the packs and standardised them all, so everyone from COLIN onward should be playing to the same set regardless of which table they're on.

It would be interesting to hear from others who attended Colin in 2023 and 2024 how much of a difference this made. For me it was huge!

Excluding finals (which used Adam's separate pack both weekends), you play or host 90 letters rounds across the Colin weekend. I counted how many 9s were available in rounds I played or hosted.
2023: Only 1 nine available across the whole weekend!
2024: 6 nines across the whole weekend.

[Adam S Latchford]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

They used to. Adam 'awesome' Beach has gone through the packs and standardised them all, so everyone from COLIN onward should be playing to the same set regardless of which table they're on.

It would be interesting to hear from others who attended Colin in 2023 and 2024 how much of a difference this made. For me it was huge!

Excluding finals (which used Adam's separate pack both weekends), you play or host 90 letters rounds across the Colin weekend. I counted how many 9s were available in rounds I played or hosted.
2023: Only 1 nine available across the whole weekend!
2024: 6 nines across the whole weekend.

Yeah I noticed lots of fruitful nines popped up. Had maybe 7 available across the weekend in games I played in (inc final) ? Also there were less of the horror show selections (although i did host a game with qbxvwiiuu, but that felt like an absolute anomaly)

[Ben Wilson]

It's worth pointing out that this is the first year that there have been 4 people scoring 400+ at COLIN, and Jack's aggregate of 448 over 6 games smashed the previous 6 game record (but was a few shy of the record average).

It's also worth pointing out that the two guys above had 2 of the 3 highest scores throughout the weekend, so it'd be good to hear from others about how they feel the games went (at least until I get the stat file updated to add a bit of data to the argument).

[Graeme Cole]

Following from this post...

If I had one small suggestion, it would be that if there are table numbers, it's best to have them explicitly labelled rather than people having to infer them from the set numbers on the lids of the letters/numbers boxes, which are generally harder to find with the lids often put upside down after being taken off.

I noticed this too. Finding the correct table was an exercise in trial and error.

Even when the table number is on a separate piece of paper, after the first round it usually ends up being moved out of the way or covered by something, especially if there isn't much space on the table.

What about blu-tacking the table number label to the side edge of the table, such that the label hangs vertically? Then it's the best of all worlds: it doesn't get obscured by anything, it can be seen from across the room rather than everyone having to approach each table to check its number, and it's not in the way of anything.

[Ben Wilson]

Following from this post...

If I had one small suggestion, it would be that if there are table numbers, it's best to have them explicitly labelled rather than people having to infer them from the set numbers on the lids of the letters/numbers boxes, which are generally harder to find with the lids often put upside down after being taken off.

I noticed this too. Finding the correct table was an exercise in trial and error.

Even when the table number is on a separate piece of paper, after the first round it usually ends up being moved out of the way or covered by something, especially if there isn't much space on the table.

What about blu-tacking the table number label to the side edge of the table, such that the label hangs vertically? Then it's the best of all worlds: it doesn't get obscured by anything, it can be seen from across the room rather than everyone having to approach each table to check its number, and it's not in the way of anything.

They were printed out for COLIN but not used as I saw the numbers on the lids of the sets and thought 'ah, this'll do'. Duly noted for future events.

[Matt Rutherford]

Always use the letters packs that Callum made. The ones taken from various copies of the board game all have very bad (and inconsistent) distributions, which make it less fun for most.

They used to. Adam 'awesome' Beach has gone through the packs and standardised them all, so everyone from COLIN onward should be playing to the same set regardless of which table they're on.

It would be interesting to hear from others who attended Colin in 2023 and 2024 how much of a difference this made. For me it was huge!

Excluding finals (which used Adam's separate pack both weekends), you play or host 90 letters rounds across the Colin weekend. I counted how many 9s were available in rounds I played or hosted.
2023: Only 1 nine available across the whole weekend!
2024: 6 nines across the whole weekend.

Beachy and I can both attest to the fact that redistributing packs is not a fun way to spend an afternoon.
-So to any event attendees, please put all things back as you find them, and
-To any event hosts, please reinforce above to event attendees

Thanks!

[Callum Todd]

Agreed, it may sound simple but manually redistributing the letters packs is a massive ballache. Please don't mess the packs up when playing with them everyone!

I also found the selections benefitted from the marvellous efforts of MSR and Beachy this weekend. I have to say I absolutely adored using Beachy's board and set at the Hangover though, first time I've had that on a table at a Lincoln-style event and it was excellent

[Gavin Chipper]

By the way, a few people at COLIN were emphasing the importance of shuffling the letters well and using e.g. four As coming out in a row as evidence of poor shuffling.

However, unless it's a new pack with letters clumped together, it shouldn't really make any difference. Shuffling is good practice anyway and certainly if anyone at your table was at the same table in the previous round, but as long as the pack has been "worn in" shuffling should make no difference to the quality of the selections.

[Adam Beach]

Following from this post...

If I had one small suggestion, it would be that if there are table numbers, it's best to have them explicitly labelled rather than people having to infer them from the set numbers on the lids of the letters/numbers boxes, which are generally harder to find with the lids often put upside down after being taken off.

I noticed this too. Finding the correct table was an exercise in trial and error.

Even when the table number is on a separate piece of paper, after the first round it usually ends up being moved out of the way or covered by something, especially if there isn't much space on the table.

What about blu-tacking the table number label to the side edge of the table, such that the label hangs vertically? Then it's the best of all worlds: it doesn't get obscured by anything, it can be seen from across the room rather than everyone having to approach each table to check its number, and it's not in the way of anything.

Thanks for bringing this up. Table numbers will be in the works this week ready for Liverpool and future events. I'll get some holders for them so that they can stand upright.

[Adam Beach]

Agreed, it may sound simple but manually redistributing the letters packs is a massive ballache. Please don't mess the packs up when playing with them everyone!

I also found the selections benefitted from the marvellous efforts of MSR and Beachy this weekend. I have to say I absolutely adored using Beachy's board and set at the Hangover though, first time I've had that on a table at a Lincoln-style event and it was excellent

Honestly, I had nothing to do with them for this event — all credit to MSaRse. I only did them for Rugby and Birmingham haha. And I adored our games. That was my top highlight of the weekend, and I got on the train home in very high spirits.

[Rhys Benjamin]

By the way, a few people at COLIN were emphasing the importance of shuffling the letters well and using e.g. four As coming out in a row as evidence of poor shuffling.

However, unless it's a new pack with letters clumped together, it shouldn't really make any difference. Shuffling is good practice anyway and certainly if anyone at your table was at the same table in the previous round, but as long as the pack has been "worn in" shuffling should make no difference to the quality of the selections.

I think this is a different issue (pure random / some manual de-clumping letter separation) which has been argued about to death.

[JackHurst]

Top tip for hosting games at Colin style co-events:

A game changer for me as an event attendee was when I started putting no effort/caring into playing along with games that I am hosting. It has many benefits:

- I'm able to prioritizing hosting effectively and make fewer mistakes with timing. I can also get the allowed words list open during the round (hidden from players!) So that I can immediately adjudicate and keep things running smoothly.
- I stop myself going into cognitive overload (because hosting a game alone is an intense cognitive load. Doing this while trying to play along isn't sustainable). This improves my overall performance! Many players seem to fatigue towards the end of an event and I think I stopped suffering from this once I stopped caring about playing along with games I'm hosting.
- It's actually quite rude to insist on telling the players how you did after every single round when you are host. You are host not player, so just let the players have their moment in the spotlight please.

[Gavin Chipper]

I don't tend to really play along as host. I'll glance at the letters and might spend a few seconds on the numbers but if I don't solve it I leave it to watch the time.

Also has this been said - as host, don't write the letters down on the scoresheet until the time starts. You don't need to interrupt the letters selection process to do something you will shortly have 30 seconds to do.

[Graeme Cole]

I don't tend to really play along as host. I'll glance at the letters and might spend a few seconds on the numbers but if I don't solve it I leave it to watch the time.

Also has this been said - as host, don't write the letters down on the scoresheet until the time starts. You don't need to interrupt the letters selection process to do something you will shortly have 30 seconds to do.

+1 to both things here.

Over many years of playing at events, I've optimised my Lincoln-style hosting procedure as follows:

• Start the timer as normal after the letters have been picked.
• Then, write down the letters on the scoresheet. If it's a numbers round, then depending on how the target was selected, it might not be displayed anywhere, so write down the target before writing the selection.
• Then, put the selection into Lexplorer or Jack's app to see the maxes, which often saves time afterwards.
• By this time there are usually about five seconds left. I tell the players there are five seconds left.
• When the time expires, take the players' declarations and words, and check/recheck Lexplorer if necessary.

All this means I have very little time, if any, to spot words myself. Sometimes, in a numbers game, the players have finished and have agreed to end the round early before I've noticed there's an easy solution, because I'm still writing the numbers and target on the scoresheet.

Other people have their own ways to fill in the scoresheet...

Lawful good: Write the selection on the scoresheet during the 30 seconds.
Lawful neutral: Write the selection while dealing out the letters.
Lawful evil: Suddenly remember to write the selection after clearing the letters away for the next round.
True neutral: Leave the selection column blank. Nobody actually needs it.
Chaotic good: Ignore the order of the letters. It's easier to remember the max and the remaining letters, so just write something like RADIALSQO.
Chaotic neutral: Don't write anything on the scoresheet until the end of the game. Assume the players are keeping score.
Chaotic evil: The labels on the various scoresheet columns are for decoration only. Write the score in the word column, the selection in the score column, the players' names in the conundrum row, and the final score on the other side of the sheet.

[JackHurst]

[*] Then, put the selection into Lexplorer or Jack's app to see the maxes, which often saves time afterwards.

Good system. Worth noting that app app has an "other words" section too, so you only need to do one search. Enter the selection. Its the offered word isn't in the maxes list, or in the alphabetised, length sorted "other words" list, you don't need to look anything else up

Also it works offline.

Giving people the hard sell. Use my the app for word checks, its the best tool

[JackHurst]

Also has this been said - as host, don't write the letters down on the scoresheet until the time starts. You don't need to interrupt the letters selection process to do something you will shortly have 30 seconds to do.

I might be considered crazy for saying this, but lets just abolish the selection column. It serves very little purpose, adds basically no value to events, and is the cause of quite a lot of stress and faff for hosts. The only thing that matters is final score and whether there was a tiebreak.

[Thomas Carey]

I'm the same hosting order as Graeme, except I usually look up the maxes (jack's thing is best for this since it gives sub maxes etc) before writing the thing down, so that I can then tab back in to the timer and put the phone visible to the players as early as possible (usually 10 seconds in or so). It also allows me contingency time just in case I mistype the selection or want to look up a definition orsomething, but in practice this doesn't really happen.

As for solving it, sometimes I'll see a decent word come up during picking/as I'm typing the thing into my phone, but you don't want to be wasting any time playing when hosting.

Having got this down, you can get through rounds very quickly, which allows you to spend some time chatting and still finish with the other tables.

[Gavin Chipper]

I tend to time on my phone (except for numbers rounds where I use my watch and use a random number generator on my phone) as I think it's helpful to have a timer the players can see. So I check for maxes etc. after the round.

Edit - This is also why it's useful to write the selections down, as someone might clear away the letters.