2007 Red Advantage

[artdutra04]

The Red Advantage thread in 2005 referred to the fact that the Red Alliance was always the higher seeded teams in the elimination rounds. Assuming the qualification rounds are 100% random*, then the average score for both alliances should be the same. But as experience tells us, they are usually not.

Due to this fact, one team may be on a certain side of the field more often than if they were completely random. Because of the way the scores work in the game, the better you are (ringer-scoring wise) the exponentially higher your scores become. Going from 1 to 2 tubes is not nearly as much as an increase as going from 5 to 6 tubes, or from 7 to 8. So it is in these phenomenal tube scoring robots that "outlier" scores happen.

Because these outlier scores can drastically alter the average, and because of the fact that due to the altered randomosity of the match generation that teams may play on one alliance more than the other, these outlier scores will often end up on one alliance more than the other.

Using data from the first 51 matches at the NJ Regional, the average score the the red alliance was just north of 33 points, while the blue alliance lagged at about 14 points. In the data used, there are several outlier scores from the red alliance (260, 98, 128 and 236 pts). If we exclude these from the average, then the average drops to only 20 and change, which is much closer to the blue alliance's average score.

But also to note is the standard deviation of the scores. Because of the outliers in the red alliance, the standard deviation for scores in the red alliance was 51.7195 points, whereas the blue alliance was only 18.676, meaning that red has a much higher spread of data. But if we again exclude the outliers, then the standard deviation of the red alliance drops to 21.087 pts, which is much closer to the blue alliance.**


* The reason matches are not completely random is because of the algorithms used to "randomize" the matches. Although it may seem like random, adding little things like a minimum time span between a single team's consecutive matches alters the randomness of the results. Now the results are no longer 100% random, but rather the sort of "equally-spaced-out-data" random that people generally associate with randomness.

** These results of course only count for the blue alliance and not The Blue Alliance, as that always tops the lists.

[amos229]

At BAE it seemed to not only be blue in the triple digits but it seemed to always be bue winning the matches.

[LocknLoad]

Red Alliance won at GSR

[BrianR]

I went through the qualification matches for data mining, and I noticed a couple of things. Unfortunately, I modified the document, and I didn't preserve the Red vs Blue scores, rather they are now Winner and Loser. In any case, when I did a quick sum of the Red scores vs. the Blue scores over the first four regionals (BAE wasn't posted), I noticed that Red had overall anywhere from slightly more to double as many points as Blue. This sure seems like a red advantage to me. The one regional where it was very clear was at VCU where Red literally demolished Blue.

Also, the other point that I found VERY interesting was that the number of teams that scored over 30 points and lost was very low. In fact, 98.515% of all losing teams had fewer than 30 points. This would seem to indicate that anyone who can score over 30 points (a 12" ramp) every match, should be well on their way to finals, if not one of the top seeds.

What do you think?

Oh yeah, this data constitutes 270 qualification matches, and none of the finals matches, as those are clearly a different caliber, and need to treated as a seperate type of match so as to prevent skewing the data.

[Barry Bonzack]

Originally Posted by artdutra04

Using data from the first 51 matches at the NJ Regional, the average score the the red alliance was just north of 33 points, while the blue alliance lagged at about 14 points. In the data used, there are several outlier scores from the red alliance (260, 98, 128 and 236 pts). If we exclude these from the average, then the average drops to only 20 and change, which is much closer to the blue alliance's average score.

But also to note is the standard deviation of the scores. Because of the outliers in the red alliance, the standard deviation for scores in the red alliance was 51.7195 points, whereas the blue alliance was only 18.676, meaning that red has a much higher spread of data. But if we again exclude the outliers, then the standard deviation of the red alliance drops to 21.087 pts, which is much closer to the blue alliance.**
[/size]

Using this information we can say with confidence that 68% of red alliance matches from this regional will have a score between 12 - 54 points. 68% of blue matches would be between 0 - 34 points.

My guess is that on a 95% scale that there is not a large enough margin to say this is significant, but I would have to have the data in front of me...And its been a couple years since I took statistics.

Also, I tend to agree with Karthik and believe that this may be a rare instance where medians are more powerful than averages because of the way the game is scored.

[aksimhal]

If all this about the red allience advantage is true, than what makes the red allience stronger?

[Wedge34]

you can see red much better from the station, and in STL, you did not have meetool birgade screaming behind you

[VanMan]

We never scored in triple digits and our opponents never scored in triple digits because we are so good at defense.

[StephLee]

At VCU, the total score for the entire competition by color breaks down like this:

Qualifiers only:
Red - 1116; Blue - 1535
Red average: 12.7
Blue average: 17.4

Eliminations only:
Red - 665; Blue - 627
Red average: 36.9
Blue average: 34.8

All matches:
Red - 1781; Blue - 2162
Red average: 16.8
Blue average: 20.4

For VCU at least, the red side didn't seem to have an advantage when the matches were supposedly random. It makes some sense for the red side to score slightly higher in the eliminations since red is the higher seeded side.