World championship eliminations alliances

Hey chief Delphi with qualification matches winding down to the end what will a world championship alliance look like this year or a division winning alliance?

I wrote a quick little script to simulate alliance selection, albeit in a relatively naive way. It assumes each team will be picking the highest available OPR team, and it assumes nobody declines. Again, this isn’t perfect, it’s just assuming each alliance wants to get the highest OPR possible.

arc alliances:
[1501 - 60 (1), 4103 - 51 (6), 180 - 41 (24), 3098 - 41 (53)],
 [4003 - 48 (2), 4564 - 49 (8), 5878 - 41 (17), 2881 - 40 (57)],
 [3166 - 37 (3), 1768 - 48 (20), 2338 - 41 (35), 1671 - 39 (39)],
 [2064 - 46 (4), 1747 - 47 (5), 5401 - 41 (50), 3847 - 39 (32)],
 [45 - 37 (7), 207 - 47 (44), 498 - 42 (72), 1986 - 39 (16)],
 [3534 - 39 (9), 2637 - 46 (21), 4039 - 42 (30), 4183 - 39 (29)],
 [1257 - 41 (10), 1310 - 45 (11), 5926 - 43 (59), 4911 - 38 (18)],
 [3932 - 38 (12), 1318 - 45 (28), 1540 - 44 (33), 1391 - 37 (19)]]
cars alliances:
[973 - 59 (1), 1024 - 67 (7), 525 - 43 (33), 2403 - 43 (29)],
 [2590 - 54 (2), 868 - 64 (42), 2655 - 43 (27), 2771 - 42 (26)],
 [225 - 57 (3), 2052 - 57 (8), 2996 - 43 (25), 3970 - 42 (45)],
 [5895 - 54 (4), 610 - 56 (12), 4362 - 43 (23), 135 - 42 (49)],
 [2122 - 56 (5), 41 - 52 (15), 1625 - 43 (53), 5907 - 42 (16)],
 [1718 - 45 (6), 2067 - 51 (37), 904 - 45 (30), 1987 - 41 (32)],
 [1918 - 47 (9), 3824 - 50 (47), 4028 - 45 (19), 4264 - 41 (13)],
 [2137 - 49 (10), 3538 - 47 (43), 2474 - 47 (21), 2486 - 41 (24)]]
carv alliances:
[133 - 64 (1), 230 - 74 (3), 5048 - 43 (55), 1086 - 43 (49)],
 [330 - 55 (2), 842 - 62 (5), 120 - 43 (62), 3452 - 42 (43)],
 [27 - 58 (4), 1114 - 55 (27), 1775 - 45 (58), 78 - 41 (18)],
 [5460 - 54 (6), 1305 - 53 (7), 359 - 46 (42), 3419 - 41 (46)],
 [2481 - 50 (8), 360 - 52 (19), 4451 - 47 (17), 1481 - 40 (48)],
 [3357 - 48 (9), 2767 - 50 (13), 1262 - 48 (12), 5924 - 40 (37)],
 [57 - 35 (10), 3478 - 50 (38), 1058 - 48 (51), 2648 - 40 (31)],
 [503 - 39 (11), 910 - 49 (45), 20 - 48 (16), 228 - 39 (34)]]
cur alliances:
[3546 - 47 (1), 166 - 60 (10), 4469 - 42 (28), 2175 - 41 (24)],
 [1089 - 45 (2), 3641 - 59 (9), 3506 - 42 (31), 3770 - 41 (32)],
 [25 - 50 (3), 5406 - 56 (5), 3234 - 42 (48), 2168 - 41 (45)],
 [5803 - 36 (4), 1885 - 56 (11), 3339 - 43 (33), 3618 - 41 (15)],
 [639 - 42 (6), 1261 - 54 (13), 2974 - 43 (29), 4180 - 40 (47)],
 [3005 - 37 (7), 694 - 53 (8), 379 - 44 (16), 1983 - 39 (40)],
 [4061 - 37 (12), 876 - 51 (21), 3990 - 46 (38), 3656 - 39 (37)],
 [836 - 31 (14), 858 - 48 (17), 2557 - 46 (22), 2883 - 39 (18)]]
gal alliances:
[195 - 70 (1), 365 - 67 (3), 1153 - 38 (13), 2502 - 37 (48)],
 [3683 - 49 (2), 558 - 63 (8), 245 - 39 (20), 68 - 36 (26)],
 [987 - 54 (4), 494 - 59 (9), 1065 - 40 (27), 4063 - 36 (29)],
 [3539 - 47 (5), 341 - 55 (6), 6166 - 41 (16), 2638 - 36 (66)],
 [5813 - 49 (7), 1124 - 55 (32), 2468 - 41 (19), 5006 - 34 (18)],
 [333 - 40 (10), 1023 - 54 (28), 3637 - 42 (33), 1701 - 34 (22)],
 [2987 - 39 (11), 2439 - 50 (21), 870 - 48 (46), 2642 - 34 (39)],
 [176 - 46 (12), 175 - 50 (14), 5448 - 49 (44), 6175 - 34 (17)]]
hop alliances:
[148 - 79 (1), 971 - 70 (4), 172 - 42 (30), 2438 - 41 (50)],
 [1425 - 49 (2), 1678 - 70 (3), 5254 - 42 (46), 2990 - 41 (35)],
 [4334 - 59 (5), 294 - 56 (7), 1684 - 42 (36), 5431 - 40 (16)],
 [303 - 34 (6), 4587 - 54 (11), 3604 - 42 (53), 4377 - 40 (38)],
 [1712 - 34 (8), 346 - 52 (9), 2013 - 42 (17), 1011 - 39 (43)],
 [231 - 32 (10), 869 - 51 (40), 70 - 43 (28), 5618 - 39 (15)],
 [1736 - 49 (12), 1538 - 49 (21), 1730 - 44 (51), 2471 - 38 (44)],
 [686 - 35 (13), 4967 - 47 (14), 1323 - 44 (25), 2046 - 38 (23)]]
new alliances:
[1241 - 62 (1), 5172 - 73 (10), 122 - 45 (37), 5505 - 45 (63)],
 [254 - 70 (2), 179 - 70 (5), 1143 - 45 (61), 2614 - 44 (39)],
 [1519 - 67 (3), 3620 - 70 (9), 3481 - 46 (44), 3284 - 44 (25)],
 [67 - 66 (4), 118 - 62 (11), 125 - 48 (33), 1640 - 43 (17)],
 [5150 - 38 (6), 4678 - 62 (21), 3476 - 50 (32), 3937 - 42 (34)],
 [4001 - 45 (7), 217 - 58 (8), 3663 - 50 (14), 2393 - 41 (36)],
 [1418 - 36 (12), 3314 - 57 (26), 3360 - 50 (47), 5124 - 41 (30)],
 [229 - 39 (13), 1731 - 55 (16), 16 - 52 (42), 708 - 40 (19)]]
tes alliances:
[2056 - 68 (1), 1690 - 69 (4), 111 - 41 (24), 74 - 41 (11)],
 [2451 - 59 (2), 33 - 62 (12), 5437 - 42 (19), 2054 - 40 (23)],
 [2834 - 51 (3), 3130 - 54 (9), 2415 - 42 (51), 329 - 40 (14)],
 [1806 - 54 (5), 5114 - 50 (38), 3245 - 43 (30), 1306 - 39 (34)],
 [2823 - 45 (6), 3238 - 49 (15), 461 - 43 (29), 2783 - 39 (32)],
 [4488 - 39 (7), 548 - 46 (48), 5030 - 43 (57), 2363 - 39 (40)],
 [4391 - 44 (8), 177 - 45 (21), 3044 - 43 (46), 4930 - 39 (49)],
 [1675 - 37 (10), 1099 - 44 (17), 4468 - 43 (16), 2522 - 39 (13)]]

Syntax is [TEAM1 - OPR1 (SEED1), TEAM2 - OPR2 (SEED2), … ]

Where the seed is where they originally finished in the division.

Source code can be found here

There’s more to alliances than just OPR. You have to take into account a team’s strategy, personal opinions, and (in my opinion) a Defense bot meta-game. At least until the teams reach Einstein.

Yeah, I’m aware… That’s why I said it was a naive simulation.

One glaring omission is 1619 from Carver. No way they don’t get picked early.

Not criticizing your effort or method, just noting some of the problems with OPR. Thanks for pulling this together as a rough gauge of what to expect.

Yep, one thing that’s particularly interesting about this is that in Carson, Team 2590 would be picking team 868, the 42nd seed in the first round. I’d say there are a few issues :wink:

I really hope people will take defense into account. We (3374) played some hardcore defense after figuring out mechanical issues and led us to the final two matches being smashing victories. Unfortunately, we ended as 43. I’d like to see how the defense role plays into alliance selection.

Archimedes alliance selection is going to be interesting to watch.

This is crazy. I never would have imagined us as a potential #1 pick in a championship subdivision! I know Curie is a relatively weak division, but wow! This weekend has been a ton of fun for us!

Seeing as 868 has been putting in upwards of 8 high goal shots in their recent matches (from the outerworks) consistently been hitting their auto goals, AND shown they can climb… I’d say you chose a bad example for “OPR issues” :wink:

Sometimes schedules are rough!

Well, I’m not saying they aren’t pick worthy. Maybe just not 2 seed first pick pick-worthy. But TBH i haven’t been watching that division. I could be very wrong.

First Round Pick (Top 4 seeds):

Best High Goal Shooters: The high goal shooters shooting on average of 6+ boulders will be picked very quickly. The alliance captain and first pick of the top 4 seeds will probably be able to get the tower down to 0 with some time to spare. Scaling on the high goal shooters makes them more desirable.

Two-Boulder Auto: Any robot performing a two-boulder auto will be picked in the top four picks. This autonomous advantage is huge. Though many alliances in eliminations will get the tower down to 0 no problem, the two ball autonomous will help if the opposing alliance plays defense.

Cheval/Portcullis Auto: Teams that can do this will be desired, especially if their alliance captain can do a two-boulder auto. This allows cycling robots to do the easier categories, allowing them to score faster (more high goals, more points, more chance of capture).

Quick Defensive Crossers: Obviously, the top teams will be able to cross Category A, B, and D by themselves, and will be able to do so very quickly. Some teams may even be able to do Category C

First Round Pick (5th to 8th seeds)

More High Goals: Despite not being the best high goal shooters, the alliance captains will choose the next best ones, as having high goals is essential to any alliance.

High-Goal Auto: Teams that can demonstrate a high goal autonomous, particularly in multiple defensive positions, will be quickly picked if available.

Most Scalers: By this point, most scaling robots have been picked. The scalers that remain are moderate in speed (10-15 secs), while the top teams can scale in 3-8 seconds. The extra 10 points is too valuable to pass up.

2nd Round Pick:

Remaining Scalers: Any scaling robot that can scale reliable in under 15 seconds will be picked, probably by the 5th-8th seed.

High-Goal Shooters?: This one is hard to estimate. Most of the high-goal shooters left will either inaccurate or unreliable. This creates a risk for this alliance, but dark horse shooters will likely be picked. High-Goal shooters in the second round will be picked more often by the 5th-8th seed.

Low-Goal Shooters: If a team can demonstrate quick low-goal shooting, while still being able to cross Category A, B, and D.

Crossing Auto: Crossing a defensive in autonomous is the minimum. If a team cannot successfully cross a defense, they will more than likely not every be chosen.

3rd Round Pick:

Reliable Robot: The 3rd pick will have to be a very reliable robot, that hasn’t experienced many mechanical issues, as this robot will be the stand-in in case one of the other three break. If this robot fails as a stand-in, the alliance will be in a very big issue playing a 2v3

Defensive Robot: Having a defensive robot will allow the alliance to substitute it in as needed to change the strategy from total offense to incorporate defensive elements.

tl;dr

Captain- High Goal Shooter
1st Pick- High Goal Shooter
2nd Pick- Quick Low Goal Shooter/Underestimated High Goal Shooter
3rd Pick- Reliable Defensive Robot

I know my team 3234 can do all or a good chunk of the defenses the only things we can’t do is shoot high and go.under the low bar I don’t see why we don’t get picked in curie we hung almost every single match we were in besides maybe were not compatible or for some reason and good luck to all team on the final day of worlds and may the best bots win

I totally agree. For many of our matches, we were with robots that were having trouble working or the other teams broke down during the match. We also ended up playing a fair bit of defense, including building a blocker. Thus, we ended up as Rank 68. I too hope that teams will not only take the rank into account when choosing teams, but also the merits of the team when paired with a strong alliance.