This is something I have been thinking about for awhile now. what would a graph look like that plotted your seeding rank vs. your chance to be on the winning alliance?
Does seeding first actually give the highest chance of winning? (and by how much?) Who has a higher chance of winning, 7th seed or 8th seed? Is seeding higher really always better, and if not, are there circumstances where throwing a match give you a higher chance of winning the regional?
Unfortunately, I don’t know how to compile the FRC/TBA data to make a graph like this. Is there anyone else who would be interested enough to do it?
This has the data you are looking for this year. If you search through the white papers or through the thread, you can probably find similar documents.
here is a quick graph. I’m going to work on seed next (not alliance, but after qualifications are over)
Courtesy of the 2834 scouting database. Qualification seeding rank on the left, number of event winning teams on right.
1 51
2 34
3 15
4 26
5 12
6 9
7 13
8 3
9 2
10 8
11 3
12 2
13 6
14 3
15 5
16 4
17 7
18 10
19 4
20 4
21 2
22 10
23 6
24 8
25 3
26 4
27 1
28 5
29 4
30 6
31 7
32 5
33 1
34 3
35 1
36 1
37 0
38 5
39 2
40 2
41 2
42 1
43 1
44 2
45 0
46 1
47 1
48 1
49 1
50 0
51 2
52 0
53 0
54 1
55 0
56 0
57 1
58 0
59 0
60 1
61 0
62 0
63 0
64 0
65 0
66 0
67 0
68 0
69 0
70 0
71 0
72 1
73 0
74 0
75 0
76 0
77 0
78 0
79 0
80 0
81 0
82 0
83 0
84 0
85 0
86 0
87 0
88 0
89 0
90 0
91 0
92 0
93 0
94 0
95 0
96 0
97 0
98 0
99 0
100 0
I have also attached a graph of the top 24 seeds.
My total number of event winners was 313, which is not divisible by 3, probably due to backup teams.
It is interesting that the number of 4th seed winners exceeds the number of 3rd seed winners by an appreciable margin. I wonder what could cause that? It might just be noise though, I’ll add in results from the past few years tomorrow.
Please dispense with all consideration of ‘throwing a match.’ You have alliance partners who very likely want to win and you owe it to them to try, even if it is not in your own team’s best interest, statistically.
If you thinking about throwing a match, don’t. Especially if your alliance partners don’t want to lose, then if your trying to throw the match and their trying to win, then they have to compete against 4 robots instead of 3. And plus one of the big points of 2014 was to encourage cooperation between teams. It is just not in the spirit of FIRST to throw a match. Secondly as an alliance captain for elims, and I see that a team threw a match, I don’t want them on my alliance because that team is just in it for themselves and not the alliance. Some food for thought.
While I was not around for the madness, I understand there was a “6v0” problem due to the way qualification worked in some previous years. When matched against a much more powerful alliance, it was to the “weaker” alliance’s advantage to score in the “better” alliance’s goal(s). Winning is best, for sure, but the rules in a particular year may make losing spectacularly more beneficial than losing by a closer margin.
On the flip side, there have been years where the seeding points were awarded based on the losing team’s score, thus incentivizing the winning alliance to not play as well, or even score for the other alliance.
Agreed. “throwing a match” is just about as far from gracious professionalism as you could possibly get.
If I knew we had lost a match because an alliance member determined it was better for them to lose, there is no word in the English dictionary to describe my disgust.
What’s really interesting are the spikes at 18 and 22. I’m guessing that those are most often the remaining picks at the end of the snake draft. Would be interesting to look at other years.
Or to look at average draft position compared to rank. Intuitively, one would assume it to be fairly linear for a while, before more or less flatlining.
Here is the data from 2011-2014. The top 8 are roughly what I had anticipated. 5th seed won more frequently than 4th seed, and 7th seed won more frequently than 6th seed, but not by any substantial margin.
What I find interesting is that any seed between ~8 to ~30 has (very) roughly the same number of event winning teams as any other seed within this range.
winning teams at different seeds 2011-2014.xlsx (33.4 KB)
winning teams at different seeds 2011-2014.xlsx (33.4 KB)
Thanks inking! This seems to answer the OP’s questions pretty completely. It’s interesting how steep the droppoff was some years with number of teams winning an event at each rank. In 2012 only 1 team won an event while ranked 5, but in 2013 there were 13. That could be noise I guess, but the games could have had something to do with it as well. One point of clarification for everybody, because of the picking among the top 8, it is likely seeds 9 and 10 will be a captain which is why 9 and 10 show a higher winning percentage than 11-15. I know other threads have analyzed alliance rank 1-8 and their chance of winning in each year, so I won’t go into that here. Just thought I’d point it out.
With the thread entitled Rank vs Blue banner chance, I would love to see a list of teams this year who won a Chairman’s Award and what their rank was, since that is also a blue banner 
I think that James makes an interesting point about rank vs. chairman award. I wonder if there is something to be said about building a robot that can play the game but then focusing on chairman award submission. Rather than a super competitive bot, a super strong chairman’s submission. I personally don’t think that that is the way to go, but is there some validity to this strategy? I don’t know.
Here are the ranks for all the Chairman’s Award winners from last season.
Rank Number
1 9
2 9
3 7
4 6
5 4
6 4
7 1
8 5
9 2
10 3
11 4
12 3
13 3
14 1
15 1
16 3
17 1
18
19 2
20 2
21 2
22 2
23 2
24
25 1
26 2
27 3
28 3
29
30 1
31
32 1
33
34 2
35
36 1
37 1
38 1
39 1
40
41 1
42
43
44 1
45
46 1
47
48 1
49
50
51
52 1
53
54 1
55
56
57 1
58
59
60
61
62
63
64
Thanks for the chairman’s rankings. I guess it seems that the top teams on the field are the top teams off the field as well. I would have expected more winners outside the top 8. But very interesting. I think answered my question. Very interesting results.
Just a programmer’s 2 cents, try using the D3.js library to visualize data from TBA. It’s very extensible and slick.
Disclaimer: When I say throwing a match would be helpful I don’t mean that you should do it, I just wanted to know if statistically there are times where you would be in a better position if you threw the match.
Now back to the thread.
Thanks inkling! That’s exactly what I was looking for. You can make some interesting observations from those graphs, for instance, I didn’t realize how much more likely a #1 seed is to win than a #2 seed. I also think that the graph does show that there is really no situation where throwing a match would be helpful.* And even if it would be, most good teams would see that you threw the match and wouldn’t pick you anyway.
*(unless maybe you are #15 and you can drop exactly 2 spots.)
For completeness, here are all blue banner recipients with their seeding position. The data are averaged from 2011-2014. Thanks to team 1114 for use of their scouting database.
There is a nice correlation between ranking high and number of Chairman’s or EI wins. The data for RAS are much more scattered. I’m not sure that the RAS graph actually says much of anything. Looking at this graph without fully understanding it might imply that it is better for rookies to seed lower at an event if they want to win RAS. However, that would be a false conclusion since rookies are generally not uniformly distributed throughout the seeds at an event.
EDIT: mislabeled axes, see next post
blue banner teams versus seed.xlsx (34.9 KB)
blue banner teams versus seed.xlsx (34.9 KB)
I mislabeled the axes on the spreadsheet from my last post.
This one makes more sense.
blue banner teams versus seed.xlsx (33.8 KB)
blue banner teams versus seed.xlsx (33.8 KB)