Reducing the number of perfect scores at an event may not decrease the number of individuals tied for some trophy.
In fact, it may well INCREASE that number.
I agree with Tim. Losing a trophy on a tie break when you’re a kid never feels good. Yes, the tie scorers scored just as well as the trophy winner.
Now for the next question…did the also-rans get “robbed” by the pairing gods by not getting harder opponents who would have produced better tie breaks?
Or did they get “blessed” by the pairing gods, who gave them easy pairings so they could win the games, earn the points and get the tie?
What you can do for the kids (and their parents) is make sure everyone understands how pairings work, and the many, many different ways one can have a good day playing over the board–ways that aren’t necessarily measured by the piece of plastic you take home. Like, maybe you picked up 50 rating points. Or your tie breaks were as good as those of the third-place winner. Or you upset someone rated 150 points higher.
At the expense of producing the best possible match-ups along the way.
As you have pointed out yourself on several occasions, before you can declare which match-ups are “best”, you must declare what the objectives of the match-ups are.
Bill Smythe
Ever wonder why we don’t use ‘seeded’ pairing? For example, in an 8 player score group you would pair:
1 vs 8
2 vs 7
3 vs 6
4 vs 5
No, why?
Because I’m the curious type. (You can take that any way you want.)
Conjecture…
That would give the middle player the bye instead of the lowest rated in case of an odd-numbered group. And it turns into a cakewalk for a GM in a large open.
The seeded match up is ideal for a knock out tournament to produce the “most exciting games at the end,” but doesn’t seem to produce competitive matches in every round of a Swiss.
The point of asking the question is to help make people aware what the purpose is of the pairings they’re generating. (IMHO, if you don’t know WHY the pairings are they way they are or the advantages and disadvantages of various pairing systems, you might as well do totally random parings.)
As noted by others, not every pairing method is going to have the same goal in mind. (That’s certainly the case for the Verber pairings.)
It’s ancient history (especially for TDs who started directing after WinTD and SwisSys came along), but before I ran my first tournament using computerized pairings I ran dozens of simulated events, trying to get a feel for what the computer did differently than I would have.
The first event I ran using the computer had an open section with about 40 players in it, including over half of the experts and Masters in Nebraska at the time.
I did that section by hand AND by the computer, and there were no differences in any pairings. I posted the computer printed pairing sheets.
Interestingly enough, several of the players commented afterwards that the pairings were fairer.
There was a famous management study done a number of years ago, where they gave a group of junior managers, middle managers and senior managers a series of reports, some hand-written, some typed, and some computer printed. The reports all had the same information in them. (Each subject saw just one of the reports in any set of 3.)
They asked them to judge the accuracy of the reports. As might be expected, the junior managers gave much higher scores to the computer printed reports than to the hand written ones, but the senior managers were the other way around.
This method would have the advantage of avoiding the pairing discontinuity in the middle of the group. With standard pairings, the player just below the middle is paired against a GM while the player just above the middle is paired against a foregone conclusion. With folded (or “seeded”) pairings, by contrast, a small difference in two players’ ratings (even near the middle) will result in a small difference in the ratings of the opponents they are paired against.
Bill Smythe