Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Rules/Strategy (http://www.chiefdelphi.com/forums/forumdisplay.php?f=6)
-   -   Alliance Selection Strategy (http://www.chiefdelphi.com/forums/showthread.php?t=115300)

themccannman 25-03-2013 17:17

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by Lil' Lavery (Post 1252489)
Not going to get into a full OPR rant, but I will briefly mention something about this anecdote. Those OPR figures would only be accurate if those robots were 100% successful in their climbing attempts.

From what I've seen it looks like OPR accounts for the "average" number of hang points that a team will get per match, if they hang for 10 points 100% of the time their OPR will be higher than if they hang for 10 points 50% of the time.

Quote:

Originally Posted by Joe Ross (Post 1252512)
What data did you use to determine it was 90% accurate?

"90% accuracy" was probably the wrong phrase for that. To elaborate, OPR ranking had a .9 correlation coefficient with robots actual offensive performance.

The main point that I've been trying to get across is that a good scouting system is irreplaceable, but if you don't have one, OPR is much better than having nothing. I think we can all agree on that.

MrForbes 25-03-2013 17:24

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by themccannman (Post 1252556)
The main point that I've been trying to get across is that a good scouting system is irreplaceable, but if you don't have one, OPR is much better than having nothing. I think we can all agree on that.

I'm thinking we might use OPR as a major element of our scouting process for Championships. We will also have students scouting the teams, but probably won't scout every team every match.

Our crude 1-sheet-of-paper-per-team scouting during the AZ regional worked fine, and we we were also able to get corroborating scouting data from a friendly team.

Bwalker 25-03-2013 19:31

Re: Alliance Selection Strategy
 
What exactly does OPR stand for and what is it?

EricH 25-03-2013 19:35

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by Bwalker (Post 1252612)
What exactly does OPR stand for and what is it?

Offensive Power Ranking.

It's an estimate of how much each team scores per match, or rather their point contribution to their alliance (on average).

The biggest problem with using solely OPR is that 1) it virtually ignores defensive players, and 2) it takes a bit of setup to run the numbers properly. (It also doesn't distinguish between how points are scored--if you're a 50-point climber and you pick a 50-point climber because they have a high OPR, you'll probably be blanked by the 50-point climber that picked a pair of shooter/defense robots that combined for 50 points.)

xSAWxBLADEx 25-03-2013 19:38

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by EricH (Post 1252614)
Offensive Power Ranking.

It's an estimate of how much each team scores per match, or rather their point contribution to their alliance (on average).

The biggest problem with using solely OPR is that 1) it virtually ignores defensive players, and 2) it takes a bit of setup to run the numbers properly. (It also doesn't distinguish between how points are scored--if you're a 50-point climber and you pick a 50-point climber because they have a high OPR, you'll probably be blanked by the 50-point climber that picked a pair of shooter/defense robots that combined for 50 points.)

That is why we use a pit scout/OPR combo.

Bwalker 25-03-2013 20:00

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by EricH (Post 1252614)
Offensive Power Ranking.

It's an estimate of how much each team scores per match, or rather their point contribution to their alliance (on average).

The biggest problem with using solely OPR is that 1) it virtually ignores defensive players, and 2) it takes a bit of setup to run the numbers properly. (It also doesn't distinguish between how points are scored--if you're a 50-point climber and you pick a 50-point climber because they have a high OPR, you'll probably be blanked by the 50-point climber that picked a pair of shooter/defense robots that combined for 50 points.)

Is that determined and managed by an individual team, or are the OPR maintained by an official source?

efoote868 25-03-2013 20:07

Re: Alliance Selection Strategy
 
Ed Law does a good job posting the results after each week of competition.
http://www.chiefdelphi.com/forums/sh...d.php?t=115380

This question has been asked many times, the search function is your friend :)
http://www.chiefdelphi.com/forums/sh...d.php?t=114980

There's an executable available by Bongle, I think the last updated version was last year, but it still works for strict OPR and predicting matches.
http://www.chiefdelphi.com/forums/sh...ad.php?t=75272

EDIT:
http://www.chiefdelphi.com/forums/sh...&postcount=280

Siri 25-03-2013 20:10

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by EricH (Post 1252614)
(It also doesn't distinguish between how points are scored--if you're a 50-point climber and you pick a 50-point climber because they have a high OPR, you'll probably be blanked by the 50-point climber that picked a pair of shooter/defense robots that combined for 50 points.)

That's why if you're a 50-point climber (or anyone else, for that matter), you should also be looking at the separate auto, teleop and climb OPRs for your prospective picks. Rather difficult to make that mistake if you do.

themccannman 25-03-2013 21:23

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by EricH (Post 1252614)
Offensive Power Ranking.
The biggest problem with using solely OPR is that 1) it virtually ignores defensive players, and 2) it takes a bit of setup to run the numbers properly. (It also doesn't distinguish between how points are scored--if you're a 50-point climber and you pick a 50-point climber because they have a high OPR, you'll probably be blanked by the 50-point climber that picked a pair of shooter/defense robots that combined for 50 points.)

It doesn't virtually ignore defensive robots, it completely does. Offensive Power Ranking is just that, a measure of a robots offensive contribution to their alliance. There is also a DPR (defensive power ranking) however it's wildly inaccurate and should be ignored. The main reason (aside form accuracy) why scouting is vastly more valuable than OPR is that it tells you why a robot performs the way it does which is absolutely necessary when considering a robot to fit a specific strategy. People should stop hating on OPR so much though, it does exactly what it's designed to do. It's not designed to replace your scouting system, so don't expect good results when you do rely on it for scouting.

EricH 25-03-2013 21:37

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by themccannman (Post 1252680)
People should stop hating on OPR so much though, it does exactly what it's designed to do. It's not designed to replace your scouting system, so don't expect good results when you do rely on it for scouting.

I have to ask, if OPR is not designed to replace a scouting system, then why did you suggest earlier that a team who was unsure of their scouting just pick based on OPR? Those two statements are not entirely compatible.

OPR does do what it's designed to do: tell you how much a team can expect to score. For alliance selection, though, I probably don't care how much my alliance can score.

What I care about is how big of a point differential my alliance can produce in my favor, which is not the same thing by any means. If that means that my really, really accurate full-court shooter has to be protected by a pair of brave little toasters, then that's what I'm going to pick (though I'll probably pick rather specialized brave little toasters, just to increase that differential). If that means that my 50-point climber needs some shooters to back it up, that's what I'm going to pick. Is there a full-court shooter somewhere in my likely road? I'm going to look at teams that either have a blocker already or that can add one quickly. This sort of thing is what OPR cannot tell you, because it quite simply is not designed to do that. But it's this sort of thing that can make or break an alliance.

Bill_B 25-03-2013 22:24

Re: Alliance Selection Strategy
 
The OPR score also cannot show a trend in a team's scoring that will show up on a multi-line scoresheet that has been recorded for that team. A dragster is going much faster nearing the finish line than near the start line. No one bothers to calculate the average speeds at the NHRA events.

Ether 25-03-2013 22:30

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by themccannman (Post 1252680)
It doesn't virtually ignore defensive robots, it completely does. Offensive Power Ranking is just that, a measure of a robots offensive contribution to their alliance.

To be clear: If a robot's excellent defense boosts the offensive output of its two alliance partners, then that defensive robot will receive credit for the increased scoring during the OPR calculation.

For anyone interested in crunching numbers, there's an interesting discussion going on over here in re analysis of OPR's match results predictive efficacy.



PVCpirate 25-03-2013 22:33

Re: Alliance Selection Strategy
 
Looks like the OP's team ended up as the 7th alliance captain and lost in 2 matches to the 2 seed and eventual finalists.

Jack_O 25-03-2013 22:34

Re: Alliance Selection Strategy
 
I have one main concern with OPR. I only had time to briefly look over how it is calculated, but what I saw troubled me. One teams OPR is influenced by their alliance partners. This influence may not be very big, but it can add up. If a team goes to an event where the teams are generally "good", then their OPR will be higher. The opposite happens when the teams are generally bad. Most OPR stats are seperated by less than one point. As a result, one match with bad partners could screw everything up.

That is just my two cents. If I am wrong, or if I misunderstood something, please let me know.

themccannman 25-03-2013 22:38

Re: Alliance Selection Strategy
 
Quote:

Originally Posted by EricH (Post 1252684)
I have to ask, if OPR is not designed to replace a scouting system, then why did you suggest earlier that a team who was unsure of their scouting just pick based on OPR? Those two statements are not entirely compatible.

OPR does do what it's designed to do: tell you how much a team can expect to score. For alliance selection, though, I probably don't care how much my alliance can score.

What I care about is how big of a point differential my alliance can produce in my favor, which is not the same thing by any means. If that means that my really, really accurate full-court shooter has to be protected by a pair of brave little toasters, then that's what I'm going to pick (though I'll probably pick rather specialized brave little toasters, just to increase that differential). If that means that my 50-point climber needs some shooters to back it up, that's what I'm going to pick. Is there a full-court shooter somewhere in my likely road? I'm going to look at teams that either have a blocker already or that can add one quickly. This sort of thing is what OPR cannot tell you, because it quite simply is not designed to do that. But it's this sort of thing that can make or break an alliance.


Because I never said that, from what he posted it sounded like they didn't have a scouting system, in which case I stand by my original statement,

Quote:

Originally Posted by themccannman (Post 1252556)
The main point that I've been trying to get across is that a good scouting system is irreplaceable, but if you don't have one, OPR is much better than having nothing.

OPR is relatively accurate for finding powerful offensive teams, and nothing else. The rest of your post seems to reiterate what I said previously,
Quote:

Originally Posted by themccannman (Post 1252680)
The main reason (aside form accuracy) why scouting is vastly more valuable than OPR is that it tells you why a robot performs the way it does which is absolutely necessary when considering a robot to fit a specific strategy. People should stop hating on OPR so much though, it does exactly what it's designed to do.

The one thing that I would like to clarify is this: if you are scouting for a robot that scores a lot of points, OPR is a relatively good measurement, if you're looking for a robot to do anything else other than score, OPR is the last thing you want to rely on.

EDIT: also Ether is correct, if a robot is playing phenomenal counter-defense it will boost their OPR.


All times are GMT -5. The time now is 17:26.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi