Tournamant Pairings in Round 1

[kal5056]

JUst a poll to see what everyone things about First Round Pairings in Tournaments

[babyshark]

I much prefer to avoid playing the guy I play regularly in the first round. Other than that, I really don't have any problem with seeding or pairing the power players. I know this is an issue that can excite some fairly strong opinions, but I mostly shrug.

Marc

[hammy]

BTW, don't get me going about 'advanced' or 'accelerated' pairings. They might work in Chess but really don't do the business in a wargames tournament. I know of at least one winner in such a tournament who said that as the event went on his games got easier. My experience of AP systems is similar.

Totally random but keep clubmates apart for the first couple of rounds is the best way IMO.

[lawrenceg]

BTW, don't get me going about 'advanced' or 'accelerated' pairings. They might work in Chess but really don't do the business in a wargames tournament. I know of at least one winner in such a tournament who said that as the event went on his games got easier. My experience of AP systems is similar.

and the problem with this is....?

[hammy]

BTW, don't get me going about 'advanced' or 'accelerated' pairings. They might work in Chess but really don't do the business in a wargames tournament. I know of at least one winner in such a tournament who said that as the event went on his games got easier. My experience of AP systems is similar.

and the problem with this is....?

Must resist......

Lawrence, I am happy to chat off line about this but in my experience there are a lot of issues with AP both actual and perceived. I really don't want to go through it all again here.

[nikgaukroger]

BTW, don't get me going about 'advanced' or 'accelerated' pairings. They might work in Chess but really don't do the business in a wargames tournament. I know of at least one winner in such a tournament who said that as the event went on his games got easier. My experience of AP systems is similar.

and the problem with this is....?

Must resist......

So why raise it?

But it is a load of cock

[hammy]

So why raise it?

Because it is there in the list of options and as my 'prefered' option isn't quite there I thought I might as well mention it in passing

[hazelbark]

The problem with pairings bsaed on ratings is, the ratings must be sound. Rating across nations or even diverse clubs are akin to which is a best team the Montreal Hockey Team, the Milan Football (Soccer) or the Boston Celtics (basketball).

Of course the answer is NONE they aren't reigate...

The other thing is teh size of player pools and rounds have a big impact too.

[agorfein1]

You've missed my preferred choice as an option (and the one we use locally):

1st round paired by best historical matchup/date.

Aaron

[plewis66]

Well, I've never had this discussion before so I'm going to stick my 'penneth in.

As a largely non-tournament player who only takes part in competition very, very rarely, I would feel uncomfortable with this idea of Advanced Pairing.

I've never come across the idea before, and it seems perverse to me. I can't even really believe it means what I think it means. So, as a low ranked player, I would be deliberately drawn against other low ranked players, thus improving my chances of a good result? And highly ranked players would be deliberately drawn against each other, limiting their chances of a good result?

As a complete outsider to competition, I have to say that just sounds plain daft, and as someone who I presume is meant to benefit from the system, I would really be unhappy about it. I want to earn good results honestly or not at all.

It's all a bit too loony-left social welfare system for me.

[hammy]

Of course the answer is NONE they aren't reigate...

Actually there is no sign of Reigate domination of the FoG tournament calendar in the UK at present.

Reigate wins are few and far between. Infact I have a feeling that Nik is the only Reigate player to have won a FoG tournament but I may well be wrong.

[lawrenceg]

The main things with to consider with pairings are:

• People don't want to travel miles and pay money to play the same people they play at home.
  
  People perceive a random pairing as fair, because the random element has an equal chance of favouring anyone.
  
  With random pairing, the ultimate winner may have gained by having easy pairings and not have been the best player, which some may perceive as unfair.
  
  If you deviate from random pairings, the pairing system should aim to give the ultimate winner the hardest pairings, and all players in with a chance of winning approximately equally hard pairings. Ideally the top few finishers should play each other at some stage so that there is a direct basis for ranking them.

Also I have heard a very good player (of DBM) say that if he played another very good player in the first round, both would play conservatively, not wanting to risk defeat and consequently being out of the running for 1st place (as normally you had to win all games to win a big DBM tournament). This meant that there would be a high risk of a draw, putting both players out of the running. If they met in the last round, they would both really go for a win, taking any necessary risks. This looks like an irrational strategy for the first round meeting, but people are irrational.

The main problem in non-random pairing is how to assess the difficulty of pairings, unless you have a good system in place for rating player skill before the tournament.

[nikgaukroger]

Also I have heard a very good player (of DBM) say that if he played another very good player in the first round, both would play conservatively, not wanting to risk defeat and consequently being out of the running for 1st place (as normally you had to win all games to win a big DBM tournament). This meant that there would be a high risk of a draw, putting both players out of the running. If they met in the last round, they would both really go for a win, taking any necessary risks. This looks like an irrational strategy for the first round meeting, but people are irrational.

Only irrational if a draw in the first round of the comp put you out of the running - you could usually stand a draw out of the games in a comp weekend and still win, but not a loss (at least this was/is the perception and I recall having exactly that conversation with Graham Evans at Warfare some years ago and Dave Handley at the Challenge last year).

[lawrenceg]

Well, I've never had this discussion before so I'm going to stick my 'penneth in.

As a largely non-tournament player who only takes part in competition very, very rarely, I would feel uncomfortable with this idea of Advanced Pairing.

I've never come across the idea before, and it seems perverse to me. I can't even really believe it means what I think it means. So, as a low ranked player, I would be deliberately drawn against other low ranked players, thus improving my chances of a good result? And highly ranked players would be deliberately drawn against each other, limiting their chances of a good result?

As a complete outsider to competition, I have to say that just sounds plain daft, and as someone who I presume is meant to benefit from the system, I would really be unhappy about it. I want to earn good results honestly or not at all.

It's all a bit too loony-left social welfare system for me.

The general idea is that a poor player who won his easy first round game would meet a good player who won his difficult first round game in the second or a later round and get knocked back down.

There are other pairing schemes (not necessarily confined to first round only), which are better. If you pair simply by scores on all the other rounds, the first round pairing doesn't make all that much difference.

[lawrenceg]

Also I have heard a very good player (of DBM) say that if he played another very good player in the first round, both would play conservatively, not wanting to risk defeat and consequently being out of the running for 1st place (as normally you had to win all games to win a big DBM tournament). This meant that there would be a high risk of a draw, putting both players out of the running. If they met in the last round, they would both really go for a win, taking any necessary risks. This looks like an irrational strategy for the first round meeting, but people are irrational.

Only irrational if a draw in the first round of the comp put you out of the running - you could usually stand a draw out of the games in a comp weekend and still win, but not a loss (at least this was/is the perception and I recall having exactly that conversation with Graham Evans at Warfare some years ago and Dave Handley at the Challenge last year).

I can only report what I remember, which was that the player in question thought his chances of winning the tournament were reduced if he met another good player in the first round, compared to meeting in the last round. THis might have been in the context of Britcon, Challenge or IWF WIC.

[GHGAustin]

In the previous Austin tournament, it was by historical opponent. That is my preferred method, which is not offered as an alternative.

In the most recent Austin tournament we paired the first round random, simply because we were running a little late and I wanted to just get the games going.

[grahambriggs]

Also I have heard a very good player (of DBM) say that if he played another very good player in the first round, both would play conservatively, not wanting to risk defeat and consequently being out of the running for 1st place (as normally you had to win all games to win a big DBM tournament). This meant that there would be a high risk of a draw, putting both players out of the running. If they met in the last round, they would both really go for a win, taking any necessary risks. This looks like an irrational strategy for the first round meeting, but people are irrational.

Only irrational if a draw in the first round of the comp put you out of the running - you could usually stand a draw out of the games in a comp weekend and still win, but not a loss (at least this was/is the perception and I recall having exactly that conversation with Graham Evans at Warfare some years ago and Dave Handley at the Challenge last year).

I can only report what I remember, which was that the player in question thought his chances of winning the tournament were reduced if he met another good player in the first round, compared to meeting in the last round. THis might have been in the context of Britcon, Challenge or IWF WIC.

In my experience it depended on the numbers in the competition. in a small field a draw in the first round was fine, as all you needed to do then was win the rest of the games. The larger the field got the more difficult that became, so more pressure on getting a result to start with. The danger with getting an exact draw in the first round was always that your round 2 opponent might well be a stodgy player with a stodgy army so dificult to get a big win.

Graham

[grahambriggs]

I much prefer random draw with the exception of avoiding clubmates. I understand why seeding gets done (more chance of the best player winning the comp) but it's always seemed unfair to me.

Accelerated pairings failed to work in practice in the comp I played it in (Melbourne IWF) All it seemed to do was to put three or four third tier players at the top of the draw in round three (when the pairings were merged together).

They weren't bad players (had won first two games) but they were promptly slaughtered if they were drawn against a top gun. This meant that round three (and perhaps round four) it was a case of which top guns got an easy game and which played a tough opponent. Hence luck of the draw seemed more important, not less.

So it seemed that the strongest players could actually have a quite easy ride for the first few rounds. I'm not really sure why the practice was such when clever people assure me that the theory is sound but that's how it seemed to work.

[nikgaukroger]

In my experience it depended on the numbers in the competition. in a small field a draw in the first round was fine, as all you needed to do then was win the rest of the games. The larger the field got the more difficult that became, so more pressure on getting a result to start with. The danger with getting an exact draw in the first round was always that your round 2 opponent might well be a stodgy player with a stodgy army so dificult to get a big win.

Graham

Worst still is ending up in the "draw zone" after 2 games - very likely to get a drawmeister thereafter and die of boredom ...

[lawrenceg]

I much prefer random draw with the exception of avoiding clubmates. I understand why seeding gets done (more chance of the best player winning the comp) but it's always seemed unfair to me.

Accelerated pairings failed to work in practice in the comp I played it in (Melbourne IWF) All it seemed to do was to put three or four third tier players at the top of the draw in round three (when the pairings were merged together).

They weren't bad players (had won first two games) but they were promptly slaughtered if they were drawn against a top gun. This meant that round three (and perhaps round four) it was a case of which top guns got an easy game and which played a tough opponent. Hence luck of the draw seemed more important, not less.

So it seemed that the strongest players could actually have a quite easy ride for the first few rounds. I'm not really sure why the practice was such when clever people assure me that the theory is sound but that's how it seemed to work.

This is no worse than a random pairing in the first round, which can also give an easy ride (and sometimes easier) to a top gun in the first few rounds. The Accelerated pairing used at the Melbourne IWF reduces the chances of it happening. THe easiest ride a top gun can get is:

Round 1: a 2nd tier player
Round 2: a 2nd tier player who beat a first tier player
Round 3: a 4th tier player who beat a 3rd tier player and a 2nd tier player.

With random you can get

Round 1: a 4th tier
Round 2: a 4th tier who beat a 4th tier
Round 3: a 4th tier who beat two 4th tier.

The problem with AP at Melbourne was the rating system used to determine the tiers had a flaw in it and put quite a few people in the wrong tier. This was fixed in time for the next IWF.