Archimedes 2014!

[Ed Law]

Originally Posted by holygrail

However, in a game that is so reliant on alliance partners, it can be a bit demoralizing.
I'm hoping that the flaws in using OPR to predict match outcomes are big enough to give us room to give us a fighting chance.

That is why I don't like this year's game. The outcome of the match depends too much on your partners.
Don't worry what the OPR predicts in terms of wins and losses. I would not completely ignore it like somebody else says. Teams that went to multiple events and have a high OPR at every event got that for a reason. It is not by luck. If the prediction is 200 to 100 against you, chances are it is 99% (I just picked a random number, not based on statistics) certain that you will lose. In that case you should try something drastic because you have nothing to lose. There is also a small chance that one of your opponents robot malfunction, so never give up. On the other hand, if many of the predicted matches that you will lose is only by less than 30 points, it is a 50/50 chance you may win. Most people that does prediction use max OPR. It does not tell the whole story. You should also use last OPR and world OPR to predict to see if it is any different. Data is never ugly. It is only how you use it that can be.

[stingray27]

Originally Posted by Ed Law

That is why I don't like this year's game. The outcome of the match depends too much on your partners.
Don't worry what the OPR predicts in terms of wins and losses. I would not completely ignore it like somebody else says. Teams that went to multiple events and have a high OPR at every event got that for a reason. It is not by luck. If the prediction is 200 to 100 against you, chances are it is 99% (I just picked a random number, not based on statistics) certain that you will lose. In that case you should try something drastic because you have nothing to lose. There is also a small chance that one of your opponents robot malfunction, so never give up. On the other hand, if many of the predicted matches that you will lose is only by less than 30 points, it is a 50/50 chance you may win. Most people that does prediction use max OPR. It does not tell the whole story. You should also use last OPR and world OPR to predict to see if it is any different. Data is never ugly. It is only how you use it that can be.

I can attest to this as well. Predictions by OPR were 91% accurate on Saturday at MSC and around an 82% accuracy overall in any regional/district. Close matches are a toss up. But I'd say above a 50 point differential and you will have to come up with a pretty crazy strategy...

[cmrnpizzo14]

Match 155 - 20,900,1939. If 20 and 900 can incorporate 1939 well then they very possibly could run 41 point cycles in under 15 seconds. Think secret sauce to the HP then HP passes to 20 in front of the goal. That has the possibility to be one of the greatest matches this year.

[PayneTrain]

I wonder what predictions look like if you change OPR this year to an adjusted number like percent of OPR rank at event.

[holygrail]

Originally Posted by cmrnpizzo14

Match 155 - 20,900,1939. If 20 and 900 can incorporate 1939 well then they very possibly could run 41 point cycles in under 15 seconds. Think secret sauce to the HP then HP passes to 20 in front of the goal. That has the possibility to be one of the greatest matches this year.

We definitely plan to incorporate into the plan. This match is going to be fun!