You sports people: methods for qualifying (for darts, in this case)

[DaveC426913]

I am organizing the darts tourney for about the fifth straight year. We play once a week. This year there are about 40 players. (10 teams of 40 players each, because eight teams of 5 players each makes for a slow game).

We want to have fair teams, and I've been pretty bad at judging skill levels, resulting in very mismatched teams that are no fun to play on (causing dropouts). I rate them from 1 (worst) to 4 (best) so that assessing a team's skill level is simply matter of averaging the team score.

I do have 10 captains who are pretty good players and who can collude with me.

Ideally, I would to have a qualifying round on the first night so that I can rate everyone. But here's the problem: There's no reason why anyone (especially the sharks) would play their best in a qualifying round. It is to their advantage to play badly, so that when they are teamed up, their team is more qualified than it would be if we had accurate numbers.

How can I get an accurate measurement of everyone's skill levels without them "gaming" the system?

This test must not last more than an hour (that's two games), but even less would be awesome.I've got two ideas so far:

1. Play as usual, but don't tell anyone it's a qualifying round. Presumably everyone will play their best by default. Have the team captains judge covertly. Problem here is that I would need to have them on teams first.

2. Be honest about it but offer some sort of prize (such as the usual rottle of bum). The problem is people who will play the long game, losing here to win at the season end. Also, thats one winner in a pool of forty.

 

[BillTre]

Seems you need longer term data under competitive conditions.
I would try using last years stats, but its not clear what you have.
It sounds like you have opinions about people's performances, in the short term (just before team choosing).
This is somewhat like a manager assembling a team. They would also use opinions about the player's based on previous year's performance(s) A new person would not have such earlier records. This would be like a rookie.

If your going with opinions, then many is probably better then few.
You could even do it like a player voted player selection.

Are there any performance statistics available?
If so you could make decisions by numbers, or a combination of opinion and numbers.

In many kinds of teams, players are also chosen for particular abilities they might have in specific tasks (like pitcher or quarterback). Baseball and football are riddled with positional differences like this.
Its not clear to me if there is anything like a clean-up hitter in your darts games or if its all just adding up individual scores.

 

[Hornbein]

Dart sandbaggers? What has this world sunk to...

 

[Office_Shredder]

Not a sports idea. But if you are bad at judging skill and the other captains are firmly on your side, try this. Have them reach rank the players in a strict ordering (if they aren't sure how to order players, random is fine). Then take the average position of each player from the 10 lists, and order the players according to that. The teams are then assigned by a snake draft selecting consecutively down the list (so no one has any discretion in their picks)This relieves you of the singular burden of talent evaluation, and also means any one person making a bad read on a player's skill can't mess up a team.

 

[DaveC426913]

Dart sandbaggers? What has this world sunk to...

I know what you mean, but seriously, think of it this way:

It's not necessarily deliberate dishonesty. When you're an amateur, it takes a lot of effort and concentration to play your best, so not playing your best is simply a matter of not trying as hard as possible.

 

[DaveC426913]

I would try using last years stats, but its not clear what you have.

Stats, no. We have an impression of how 90% of the players did in previous years, but its not tracked.

It's just that it can go astonishingly badly.

Its not clear to me if there is anything like a clean-up hitter in your darts games or if its all just adding up individual scores.

No, we make no distinctions for type.
And scores are at the team level, not the player level.

 

[hutchphd]

If you have a complete list of players how about the following:

1. You choose (or assign someone to choose) the 10 captains
2. On opening night, before the first games, each captain chooses the players in three rounds of selection (order within round determined by draw for each round)
3. Perhaps each team allowed one trade at end if both captains involved agree (?)

If you want to have "playground shame", the process could be public!
Let the mayhem begin.

 

[DaveC426913]

Not a sports idea. But if you are bad at judging skill and the other captains are firmly on your side, try this. Have them reach rank the players in a strict ordering (if they aren't sure how to order players, random is fine). Then take the average position of each player from the 10 lists, and order the players according to that. The teams are then assigned by a snake draft selecting consecutively down the list (so no one has any discretion in their picks)This relieves you of the singular burden of talent evaluation, and also means any one person making a bad read on a player's skill can't mess up a team.

Awesome. Now let's try that again but this time let's pretend - for fun - that I have no idea what any of those terms mean.

 

[DaveC426913]

If you have a complete list of players how about the following:

1. You choose (or assign someone to choose) the 10 captains

Generally, ahead of time, the most skilled players have been asked, and agree, to be Captains.

1. On opening night, before the first games, each captain chooses the players in three rounds of selection (order within round determined by draw for each round)

But that just puts the onus on the captains to know how skilled all the players are, which is kind of worse. Also, it tends to mean the teams stay similar year after year.

If you want to have "playground shame", the process could be public!

Indeed. This is the very thing we want to avoid, and the primary reason why we have been building the teams privately. It is no conducive to "an evening of drinking and fun for all" if we single out bad or new players.

 

[Office_Shredder]

Let's suppose you have 2 captains and 6 players (not including the captains) for 2 teams of 4

The other players are a,b,c,d,e,f. Captain 1 ranks them

a,f,e,b,c,d

a is best, d is worst.

Captain 2 ranks them

f,a,e,d,b,c

a is position 1 in one list and position 2 in another for an average ranking of 1.5. Similarly b is 4.5, c is 5.5, d is 5, e is 3 and f is 1.5. We now order them from best to worst using this measurement. a and f are tied, so their order doesn't matter between each other, I arbitrarily put a first.
a,f,e,b,d,c

Now we do a snake draft. Team 1 picks first, then team 2, then we flip the order for the next round. a-> 1, f-> 2, e-> 2, b-> 1, d-> 1, c-> 2.

And thus your teams are made. You can assign the captains to the teams randomly - you could just put them in the ranking, but then you might get two captains on one team which I assume you are trying to avoid.

 

[hutchphd]

But that just puts the onus on the captains to know how skilled all the players are, which is kind of worse. Also, it tends to mean the teams stay similar year after year.

Well someone has to make judgments if the process is not just randomized. Why is this worse?
If you chose different captains would the mix not change?
Are Canadians perhaps too nice for blood sport?...

 

[DaveC426913]

Well someone has to make judgments if the process is not just randomized. Why is this worse?

The solution I was hoping for was heavy on data, less so on judgement. As I mentioned, some sort of quick qualifying round; I just couldn't figure out the deets.

Are Canadians perhaps too nice for blood sport?...

Back story:

We are a yacht club. (OK, actually, were a drinking club, but with boats). Darts is our off-season activity, to keep us from going stir-crazy for the six month off-season.

The primary goal is to attract new club members and to get to know each other and have fun, but do it with the premise of a purpose.

So, we want to make it as friendly and non-threatening as possible - but we still want fair teams or it stops being fun for most people.

 

[DaveC426913]

I still don't know what a "snake draft" is, but that's OK.

I think having the Captains weigh-in will work OK. They've been around long enough to have some insight, but importantly, it is not necessary for every captain to assign a level for every player. I'll simply send a spreadsheet of all players and ask them to assign levels for players they're confident they know. That should still provide enough data.