Talk Tennis

Talk Tennis (http://tt.tennis-warehouse.com/index.php)
-   Adult League & Tournament Talk (http://tt.tennis-warehouse.com/forumdisplay.php?f=35)
-   -   Playoff strategy "tournament"? (http://tt.tennis-warehouse.com/showthread.php?t=456144)

spot 02-28-2013 07:06 AM

Playoff strategy "tournament"?
 
So there has been some discussion in another thread of how much it makes sense to balance lineups vs concentrating firepower so I thought it might be fun for people to test out their theories a bit. This could fall flat or it could be interesting- we might as well try.

Lets set up a tournament bracket on the board where the rules are simple. We will have people sign up and create a bracket. There will be 5 lines of doubles to make it easy. You will have a roster of 15 players. Your goal every week is to take 3 of the 5 lines. If people are interested in this then and it works out we will try other scenarios. (Teams that are imbalanced, teams where you don't know what availability the other team has, etc)

The way it would work would be that you PM me a "lineup" of your 5 lines with your pairings. Player 15 is your best player, player 1 is your worst player. We will do this USTA style where there are no sandbagging rules.

So as an example you might send me

line 1- 15 + 14 = 29
line 2- 9 + 13 = 22
line 3- 6 + 7 = 13
line 4- 8 + 12 = 20
line 5- 10 + 11 = 21

I will compare you to your opponent and post the winner. Highest point value at any line wins. If there is a tie then I will flip a coin. I think that other tiebreakers would be too controversial... say 10 and 9 face 12 and 7. Personally I'd put my money on 12 and 7 but other people would bet on 10 and 9 so for now we will just do a coinflip.

I will post all the results between rounds so it would be similar to USTA where you can look up results of teams. Then you will send me your next round lineup and we will repeat. In this one teams are exactly even.

Anyway- this very well may be too nerdy and only of interest to me and I can accept that. But if you would like to participate post here and we will see how it goes. I just think it would be interesting to see how different strategies play out across different rounds.

LuckyR 02-28-2013 07:19 AM

If you are trying to mimic the USTA leagues, you should call the 5 lines: yellow, blue, green, red and orange and not give them numerical labels.

spot 02-28-2013 07:28 AM

If some captains want to win line 1 so the rest of their team is inspired that is their decision. The lines will be played in order where the first team to take 3 points wins and if you put your best players at lines 4 and 5 then you take the risk that they won't matter. (yes... I know... but this is a rationale I have heard many a captain use. I've seriously thought about doing this with cards and flipping them over 1 pile at a time just to make that exact point)

OrangePower 02-28-2013 08:26 AM

Interesting idea but I already have my hands full just trying to manage my actual USTA team, let alone another fantasy team :)

schmke 02-28-2013 08:46 AM

What might be interesting is to just run a simulation of equal teams with different strategies to see what happens.

damazing 02-28-2013 08:52 AM

If we had the DNTRP we could look at real matches and see if the combination of players played on different lines if the end result would have been materially different.

I've played enough in my local league to know the better and worse players and looking at the match ups last year there were all kinds of strategies that the captains used to try to get the team win. Some played their strongest pairs at D2 and D3 basically giving up D1 and others randomly putting their best pair at all of the different doubles lines throughout the season.

The teams without the depth of talent but with crafty captains definitely did better than expected by rotating their stronger pairings.

spot 02-28-2013 08:53 AM

Quote:

Originally Posted by schmke (Post 7241886)
What might be interesting is to just run a simulation of equal teams with different strategies to see what happens.

I don't think that simulation would be any more valuable than running a tournament of paper, rock scissors. There is no strategy that would beat all others, only ones that would work best against what other people choose to do. This is a classic game theory problem.

If several people wanted to PM their idea for a lineup then I would be happy to run their lineups against all other people that submitted them to show this. But in the end it simply would depend on what you think that your opponent would choose to do and what their biases are.

In the classic game theory version of this game you would simply have 100 units that you could place across the 5 lines however you wish. so you could do

96 1 1 1 1

some people think you woudl go balanced with

20 20 20 20 20

Of course that loses 1-4 against a team that dumps a line with

0 25 25 25 25

which always loses to a team that dumps 2 lines

0 0 33 33 34

which almost always loses to a team that puts out 2 great lines and 3 mediocre ones.

5 5 5 42 43

Which always loses to the balanced team of

20 20 20 20 20


And of course this is the situation with equal numbers. THink about the situation where one team had 80 points to play with and the other team had 100. The strategies change a great deal.

schmke 02-28-2013 10:36 AM

Quote:

Originally Posted by spot (Post 7241902)
I don't think that simulation would be any more valuable than running a tournament of paper, rock scissors. There is no strategy that would beat all others, only ones that would work best against what other people choose to do. This is a classic game theory problem.

If all the teams are equal, you are right, over the long haul things will average out. But the issue being debated is the effect of playing only one's stronger players vs trying to get playing time for all players on the roster, even the weaker ones. And since you seem to have given a roster of 15, this question seems pertinent.

I certainly wouldn't expect a strategy that aims to get equal playing time to work the best against one that runs out the 8 strongest players all the time, regardless of what courts are played. The question is, how much of a deficit does this put the "equal playing time" team at and does it affect team won/lost or just individual court won/lost? This is the simulation that would be interesting to look at.

ABtennis 02-28-2013 04:41 PM

Quote:

Originally Posted by spot (Post 7241684)
So there has been some discussion in another thread of how much it makes sense to balance lineups vs concentrating firepower so I thought it might be fun for people to test out their theories a bit. This could fall flat or it could be interesting- we might as well try.

Lets set up a tournament bracket on the board where the rules are simple. We will have people sign up and create a bracket. There will be 5 lines of doubles to make it easy. You will have a roster of 15 players. Your goal every week is to take 3 of the 5 lines. If people are interested in this then and it works out we will try other scenarios. (Teams that are imbalanced, teams where you don't know what availability the other team has, etc)

The way it would work would be that you PM me a "lineup" of your 5 lines with your pairings. Player 15 is your best player, player 1 is your worst player. We will do this USTA style where there are no sandbagging rules.

So as an example you might send me

line 1- 15 + 14 = 29
line 2- 9 + 13 = 22
line 3- 6 + 7 = 13
line 4- 8 + 12 = 20
line 5- 10 + 11 = 21

I will compare you to your opponent and post the winner. Highest point value at any line wins. If there is a tie then I will flip a coin. I think that other tiebreakers would be too controversial... say 10 and 9 face 12 and 7. Personally I'd put my money on 12 and 7 but other people would bet on 10 and 9 so for now we will just do a coinflip.

I will post all the results between rounds so it would be similar to USTA where you can look up results of teams. Then you will send me your next round lineup and we will repeat. In this one teams are exactly even.

Anyway- this very well may be too nerdy and only of interest to me and I can accept that. But if you would like to participate post here and we will see how it goes. I just think it would be interesting to see how different strategies play out across different rounds.

Good luck trying to figure out the strength rating of 15 players. Perhaps the strongest and weakest players will be evident, but most other players will be too close to rank. My advice, don't waste anymore time on this futile exercise.

Cindysphinx 02-28-2013 04:49 PM

Nerdy guys are Dead Sexy.


All times are GMT -8. The time now is 03:47 PM.

Powered by vBulletin® Version 3.6.9
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
© 2006 - Tennis Warehouse