c4countdown

Afternoon all

Every time I organise a tournament I always get comments from people afterwards about they prefer one format over another, or how a certain format can be improved etc etc. So I thought I'd have a poll to settle it once and for all. Pick your favourite format. Here's a quick glossary to help you-

Edinburgh style- played three-to-a-table with one player 'hosting' and the other two playing, everyone plays everyone else on the table once and hosts once. (like at COLIN and Co:Lon)

Bristol style- everybody plays in pairs and plays to the same set of letters (like at CoBris and CoNot).

'grand final' and 'knockout stage' kind of speak for themselves. The lack of either of these means winners will be determined by the games they played Edinburgh- or Bristol-style.

I've only been to one COLIN and one CONOT, and if anything I probably preferred the COLIN version, so Edinburgh-style.

One problem with Bristol-style is that whoever is arbitrarily picked to choose the numbers can have a big effect on the game. And if you're sitting at the back with people in front of you, sometimes it's difficult to get the letters down.

How do conundrums work if you're doing it Bristol style?

Michael Wallace wrote:How do conundrums work if you're doing it Bristol style?

The screen at the front displays the conundrum and you "tap" the table when you think you've got it to notify your other partner. (S)he is still in play in case you get it wrong. You should write down when you got it and what you got.

Wow. That sounds really shit.

Michael Wallace wrote:Wow. That sounds really shit.

It does, I've never played those rules but it sounds very open to abuse.

Michael Wallace wrote:doing it Bristol style [...] That sounds really shit.

Fnerr.

JimBentley wrote:

Michael Wallace wrote:Wow. That sounds really shit.

It does, I've never played those rules but it sounds very open to abuse.

Didn't Kirk claim shenanigans against Robbo at the last CoLei or something?

I call the rules CoGood and CoJeff.

I went for Edinburgh style, but I like 'em both. Long may they both live, despite their respective failings.

Jon Corby wrote:

JimBentley wrote:

Michael Wallace wrote:Wow. That sounds really shit.

It does, I've never played those rules but it sounds very open to abuse.

Didn't Kirk claim shenanigans against Robbo at the last CoLei or something?

The CoLei before that - but I wasn't watching. I was looking at trying to get the conundrum against Amie.

Rhys Benjamin wrote:The CoLei before that - but I wasn't watching. I was looking at trying to get the conundrum against Amie.

I bet that's not all you were looking at.

Matt Morrison wrote:I call the rules CoGood and CoJeff.

lol.

The problem with the Bristol format IMO is that, while you theoretically get to play more games, the whole thing gets bogged down with the overheads of looking up 30 different phonies every round etc and ends up being slower. And giving everyone the same conundrum is just a terrible idea.

I think Edinburgh with grand final - in the knockout stage, what do the knocked out ones do?

Charlie Reams wrote:

Matt Morrison wrote:I call the rules CoGood and CoJeff.

lol.

The problem with the Bristol format IMO is that, while you theoretically get to play more games, the whole thing gets bogged down with the overheads of looking up 30 different phonies every round etc and ends up being slower. And giving everyone the same conundrum is just a terrible idea.

So how does word checking work at Co:Lon? Is there a dictionary on every table, or does one central person have a dictionary (or laptop, or whatever), the bottleneck effect being reduced by the fact that everyone's rounds tend to end at different times?

Graeme Cole wrote:So how does word checking work at Co:Lon? Is there a dictionary on every table, or does one central person have a dictionary (or laptop, or whatever), the bottleneck effect being reduced by the fact that everyone's rounds tend to end at different times?

There's a central dictionary and a laptop, assuming that everything's the same as last time.

James Robinson wrote:

Graeme Cole wrote:So how does word checking work at Co:Lon? Is there a dictionary on every table, or does one central person have a dictionary (or laptop, or whatever), the bottleneck effect being reduced by the fact that everyone's rounds tend to end at different times?

There's a central dictionary and a laptop, assuming that everything's the same as last time.

Yep. I think there's only going to be one laptop this year, but you just type in the word and it tells you if it's valid or not. There will be an ODE2r as a backup if the computer breaks, but hopefully that's not going to happen...

Matt Morrison wrote:I call the rules CoGood and CoJeff.

Sigh!!! Was that necessary?

Anyway, isn't it CoBen1 and CoBen2?

Karen Pearson wrote:

Matt Morrison wrote:I call the rules CoGood and CoJeff.

Sigh!!! Was that necessary?

Anyway, isn't it CoBen1 and CoBen2?

I'm pretty sure it's CoJerryHumphries actually.

Jon O'Neill wrote:

Karen Pearson wrote:

Matt Morrison wrote:I call the rules CoGood and CoJeff.

Sigh!!! Was that necessary?

Anyway, isn't it CoBen1 and CoBen2?

I'm pretty sure it's CoJerryHumphries actually.

If we're going to be really picky it's technically CoNitaMarr and CoRexClegg.

I'm really more interested in getting a cheap quick laugh than *seriously* laying into Jeff, as I'm sure you all really know.
But to talk of what's not necessary, the dude banned me from his loosely Countdown-themed circle jerks when I didn't say a single bad thing about him, so, and strike me down, I do quite enjoy taking the piss out of him now.

I prefer the Edinburgh style, simply because I like to be able to pick my own letters and numbers. In the Bristol style, you're at the mercy of someone else's picking over 90% of the time, which denies you the tactical element of picking potentially high- or low-scoring selections depending on the scoreline in your game. Little would frustrate me more, than being 29 points down after round 6 of a 9-rounder, to have the first 8 letters of round 7 come out APTEROUS, and then have a third party pick a fifth vowel for their ninth letter.

I don't like the idea of there being a top 2 or top 4 play-off after the Swiss stages, but since any such final would never involve me anyway (other than as a spectator), my argument here is somewhat theoretical.

Matt Bayfield wrote: I don't like the idea of there being a top 2 or top 4 play-off after the Swiss stages, but since any such final would never involve me anyway (other than as a spectator), my argument here is somewhat theoretical.

Weren't you in the CoLei final once, or have I gone mad?

I like the Bristol format. There's more of a communal atmoshpere to it. I might go the whole day without talking to some people under the Edinburgh style, but at Leicester or Nottingham everyone's involved with shouting out answers and can chat more easily between rounds. I like having a grand final to end the day as well.

Thomas Carey wrote:I think Edinburgh with grand final - in the knockout stage, what do the knocked out ones do?

Last time I went to Leicester the people who got knocked out in the quarters played each other to determine places 5 to 8. Beaten semi finalists had a third place playoff.

Martin: I was referring to Swiss style formats, which I believe have less potential for shock results. Because, yes, I did unexpectedly make the final of CoLei3 [of which I'm mighty proud], though the first three rounds of the day were drawn randomly, so I only had to win two knockout games against "top" players to get to the final. Now whilst it wouldn't be completely out of the question I might make the top 2 again in a true Swiss style format, I'd have to beat 3 or 4 top players. I can't really see that ever happening.

Thomas Carey wrote:I think Edinburgh with grand final - in the knockout stage, what do the knocked out ones do?

I think that's a good point. A grand final is one thing, but anything more than that would be too much for knocked out players. I'm not sure a final is even needed.

Yeh, I'm not really a fan of randomly picked tournament games anywhere, the people you get drawn against basically count for more than skill sometimes.

Matt Bayfield wrote:Martin: I was referring to Swiss style formats, which I believe have less potential for shock results. Because, yes, I did unexpectedly make the final of CoLei3 [of which I'm mighty proud], though the first three rounds of the day were drawn randomly, so I only had to win two knockout games against "top" players to get to the final. Now whilst it wouldn't be completely out of the question I might make the top 2 again in a true Swiss style format, I'd have to beat 3 or 4 top players. I can't really see that ever happening.

But you don't necessarily have to win every game. If you lose early, you might find yourself against weaker players and win the rest and sneak in through the back door.

Joseph Krol wrote:Yeh, I'm not really a fan of randomly picked tournament games anywhere, the people you get drawn against basically count for more than skill sometimes.

The same problem exists with Swiss pairs.

Gavin Chipper wrote:

Matt Bayfield wrote:Martin: I was referring to Swiss style formats, which I believe have less potential for shock results. Because, yes, I did unexpectedly make the final of CoLei3 [of which I'm mighty proud], though the first three rounds of the day were drawn randomly, so I only had to win two knockout games against "top" players to get to the final. Now whilst it wouldn't be completely out of the question I might make the top 2 again in a true Swiss style format, I'd have to beat 3 or 4 top players. I can't really see that ever happening.

But you don't necessarily have to win every game. If you lose early, you might find yourself against weaker players and win the rest and sneak in through the back door.

Joseph Krol wrote:Yeh, I'm not really a fan of randomly picked tournament games anywhere, the people you get drawn against basically count for more than skill sometimes.

The same problem exists with Swiss pairs.

Unfortunately the only real solution to that problem with swiss pairs is 'play more games'. And it's often a struggle to fit in six as it is.

Matt Bayfield wrote:I don't like the idea of there being a top 2 or top 4 play-off after the Swiss stages, but since any such final would never involve me anyway, my argument here is somewhat theoretical.

Before Gev gets in here, I'd like to point out that at no point in this thread, did I ever say the above. In hindsight, very embarrassing!

I have a confession to make.

I have never played Edinburgh.

I decided that here is the best place to post this rather than starting a new thread.

Basically, I was thinking about the Swiss pairs/triples system, and while I think it can be a good system, I don't think it necessarily needs to be used for every tournament ever. (It's been used at all the CO events I've been to, I think). The idea is that it gives people opponents who are generally of the same standard, but this can possibly be at the expense of "fairness". It also seems odd that the first round is always between three spaced-out seeds rather than having 1, 2 and 3 on the same table etc., so the philosophy of the first round contradicts the philosophy of the later rounds.

The fairest system would probably be one that gives each player opponents that best reflect the whole opposition on offer. So, if you have six opponents, you might have one from the top sixth, one from the next sixth and so on. And I don't think people would have a problem with this. I think some of the less brilliant players like the first round of the CO event because it's the only chance they'll get to play a top player. The point is I don't think people necessarily want to just play people at their level.

A random draw (except with something to stop someone having the same opponent twice) would obviously approximate to this anyway and probably wouldn't be too bad with six games to play.

So in summary, I'd say that the Swiss pairs/triples system is a good system in its own right, but I think it would add variety if CO events didn't all use this format. And to head off the obvious comment from Mr Reams, maybe I will.

Aye.

Gavin Chipper wrote:The fairest system would probably be one that gives each player opponents that best reflect the whole opposition on offer. So, if you have six opponents, you might have one from the top sixth, one from the next sixth and so on. And I don't think people would have a problem with this. I think some of the less brilliant players like the first round of the CO event because it's the only chance they'll get to play a top player. The point is I don't think people necessarily want to just play people at their level.

I think the trouble with this (where this = the question of formats, not your specific idea) is that there are arguably two components (maybe more?) to what makes the best system. One is that it is desirable to return the best player as the winner, the other is for the players to have fun. The latter I would say is more important, but also relates to the former (e.g. I assume at least some of the better players would be a bit miffed at a system where their chances of winning get screwed because of who they happen to draw for their top sixth player). Moreover, what people would find fun is hard to quantify. For example, my intuition would be that at a tournament people want competitive games, which is clearly contrary to your view, so presumably some surveying would be good.

Taking your 'reflect the opposition' idea, I'd worry it places too much weight on rankings that are calculated prior to the tournament. Aptoranks are fine for the most regular players, but obviously get a bit dodgy for less frequent or unknown players, and if the entire structure was determined by one's ranking it has the potential to get a bit silly. I also wonder if you'd need a slightly more nuanced approach that takes into account the distribution of player ranks, to avoid problems where (e.g.) the top sixth of players includes one guy who is comparatively rubbish.

Michael Wallace wrote:

Gavin Chipper wrote:The fairest system would probably be one that gives each player opponents that best reflect the whole opposition on offer. So, if you have six opponents, you might have one from the top sixth, one from the next sixth and so on. And I don't think people would have a problem with this. I think some of the less brilliant players like the first round of the CO event because it's the only chance they'll get to play a top player. The point is I don't think people necessarily want to just play people at their level.

I think the trouble with this (where this = the question of formats, not your specific idea) is that there are arguably two components (maybe more?) to what makes the best system. One is that it is desirable to return the best player as the winner, the other is for the players to have fun. The latter I would say is more important, but also relates to the former (e.g. I assume at least some of the better players would be a bit miffed at a system where their chances of winning get screwed because of who they happen to draw for their top sixth player). Moreover, what people would find fun is hard to quantify. For example, my intuition would be that at a tournament people want competitive games, which is clearly contrary to your view, so presumably some surveying would be good.

Taking your 'reflect the opposition' idea, I'd worry it places too much weight on rankings that are calculated prior to the tournament. Aptoranks are fine for the most regular players, but obviously get a bit dodgy for less frequent or unknown players, and if the entire structure was determined by one's ranking it has the potential to get a bit silly. I also wonder if you'd need a slightly more nuanced approach that takes into account the distribution of player ranks, to avoid problems where (e.g.) the top sixth of players includes one guy who is comparatively rubbish.

Yep. I'm not sure how many tournaments have been run to different systems, and how people felt about them when they went away. And also, it's not that I have a problem with the Swiss triples system, but that it would be interesting to see a few other systems out there as well.

I think if I was doing a system on rankings, it could use Aptoranks/some other prior knowledge for round 1, but then recalculate the ranks based on the tournament once it's started. So only the round 1 games would be pre-determined. The first round could even be random.

Particularly with the moveable rankings, it won't be possible for everyone to play someone in each sixth of the rankings anyway - that was just a general goal for the system to be aiming at. I also think it would add too much complication to also look at the skill distribution. I would have thought that there must be a system out there like this anyway.

I had crafted a long, turgid reply to this until I realised that Gev's effectively backpedalled from his 'I prefer random draws' stance anyway. The system used at COLIN- one player from each of the three tiers in the first round draw- will stay. The potential for upsets is certainly there, and it beats the potential of having, say, Innis, Kirk & Ed on the same table in round 1, one of them losing both games and rampaging through the rest of the tournament. This may be elitist but I really can't see a better way of sorting it when at a tournament you only play six games. If I expanded to eight games, then some randomisation in the first round may be fun, but ask yourself this- would you rather have a fourth round of games or another four rounds of drinks?

Gavin Chipper wrote:I think if I was doing a system on rankings, it could use Aptoranks/some other prior knowledge for round 1, but then recalculate the ranks based on the tournament once it's started. So only the round 1 games would be pre-determined. The first round could even be random.

Particularly with the moveable rankings, it won't be possible for everyone to play someone in each sixth of the rankings anyway - that was just a general goal for the system to be aiming at. I also think it would add too much complication to also look at the skill distribution. I would have thought that there must be a system out there like this anyway.

Would 'recalculate the ranks based on the tournament once it's started' mean all previous knowledge was ignored? e.g. if A and B are really good with A narrowly beating B in a relatively low scoring game, whilst C and D are relatively rubbish, but C is a bit better so thrashes D, then C would be ranked above A? (Btw, contrary to how this probably seems, I'm not just being a lamer and trying to pick holes in everything for the sake of it, I think this is potentially quite an interesting puzzle but I don't quite understand the mechanics of what you're proposing.)

Michael Wallace wrote:

Gavin Chipper wrote:I think if I was doing a system on rankings, it could use Aptoranks/some other prior knowledge for round 1, but then recalculate the ranks based on the tournament once it's started. So only the round 1 games would be pre-determined. The first round could even be random.

Particularly with the moveable rankings, it won't be possible for everyone to play someone in each sixth of the rankings anyway - that was just a general goal for the system to be aiming at. I also think it would add too much complication to also look at the skill distribution. I would have thought that there must be a system out there like this anyway.

Would 'recalculate the ranks based on the tournament once it's started' mean all previous knowledge was ignored? e.g. if A and B are really good with A narrowly beating B in a relatively low scoring game, whilst C and D are relatively rubbish, but C is a bit better so thrashes D, then C would be ranked above A? (Btw, contrary to how this probably seems, I'm not just being a lamer and trying to pick holes in everything for the sake of it, I think this is potentially quite an interesting puzzle but I don't quite understand the mechanics of what you're proposing.)

I was thinking about ignoring previous ranks, as I think the Swiss pairs/triples does as well. Once the tournament has started, everything is based on the tournament. So it would mean that ranks would be all over the place for a bit, but hopefully after two rounds (four games with the triples system), things should be mostly in some sort of reasonable order, and the games for round three would be based on the ranks of players that people have played based on their current ranks.

For example, if I played ranks 1 and 2 in the first round (this probably wouldn't actually happen) but they went on to lose three and four of their first four games, their new ranks might be 21 and 36 (or whatever) and my last round draw would be based on having already played ranks 21 and 36, not 1 and 2.