The position of cable protector on the field disadvantages the red alliance in two main ways.
First, it makes the ball return exit aligned with it (which is supposed to be a red-favored exit) not as favorable, because half of the balls coming out of that exit end up on the far side of the bump. This introduces extra effort (whether due to difficulty in maneuvering, needing to slow down, or just plain annoyance) in retrieving balls that should have been easy to get.
Second, many teams prefer not to score near the cable protector due to issues with turning or bumpiness that can affect shot consistency. The cable protector creates a dead zone for shooting that extends solely into red’s half of the field (indicated by the magenta line), giving the red alliance less area of accessible field space for scoring than the blue alliance.
These issues seem minor but may have a large impact at the highest levels of play when ball accessibility/starvation and defense matter a lot. Personally, my team has found playing the left side of the field on the red alliance to feel really cramped, and there’s not a lot of room to maneuver and get clean shots off.
What are your thoughts? Is this why 14/18 matches on Einstein were won by the blue alliance? Should FIRST have put a dummy cable protector on the opposite side of the field?
I was like pretty sure they did until I saw the field for the first time. It seems like it goes without saying. iirc 2019 had a symmetrical cable bump?
We played (3) Elimination Brackets on the Red Side of the Field and generally avoided the Left Alley whenever possible, especially earlier in the season.
Whether or not this artificially biases the entire game dynamic is hard to say, since the inverse point is also true when looking from a Blue Alliance Perspective.
When playing as the Blue Alliance, it is especially difficult to play in front of the Red Hangar, since the field is physically tighter there and it’s much, much easier to find yourself either beached, or at least moving in a non-intended fashion due to the cable protector. Knowing all that we know now about how easy it is to accidentally award a Traverse Climb to a non-traverse robot, this could be more detrimental in the long run.
If anything, the cable protector could be a net wash. Sure, it makes playing offense in that specific area problematic for red, but it should also make playing defense and/or retrieving balls from the Red Tarmac and Hangar equally problematic for blue?
This is an interesting point. My team’s programmers have been able to make a 3 ball auto for when we are on the blue alliance. It is a standard 2 ball with a path that picks up a ball on the red side of the field using flappers on our intake to suck the ball in. With this, we would not be touching it where it was initially placed. While we never got to use this because of connectivity issues (a wire coming loose), we got it approved by a head ref at worlds. This would have allowed for a 42-point auto when we were on the blue alliance, but the cable protector makes it, if not impossible, very hard to do when on the red alliance.
Ignoring ties, there have been 11872 official qual matches played in 2022. Of those, red has won 5920 of them and blue has won 5952, or 49.87% and 50.13% respectively. That’s about as close to even as you can get. And just for good measure, if you run a chi-squared test on the data you get a p-value of 0.769 that the probabilities are even; not even remotely close to rejecting the null.
Code
import tbapy
from scipy import stats
tba = tbapy.TBA('')
winners = [0,0,0]
events = tba.events(year=2022, simple=True)
for event in events:
if event["event_type"] > 5: continue
print(event["key"])
matches = tba.event_matches(event=event["key"], simple=True)
for match in matches:
if match["comp_level"] != "qm": continue
winners[{"red":0, "blue":1, "":2}[match["winning_alliance"]]] += 1
winners.pop()
total = sum(winners)
print(winners)
print(total)
results = stats.chisquare(winners)
print(results.pvalue)
another factor that on some level counteracts the blue advantage is the tendency of the cable protector to divide the cargo, with the smaller arear on the red side especially around the hanger there’s a tendency for cargo to pile up in an advantageous position for red when your playing with a lot of cargo on the ground.
If this theory holds true I think we could see a clear divide based on the level of play
For playoffs you would naturally expect red to win more often because the higher alliance is usually in red. I suppose you could compare the playoff percents to a “fair” game and try to find a difference. But seeing as I’m about to board a plane that’s more work than I’m going to do right now. I posted the python code I used if anyone wants to play around with it more.
Yeah, forgot about that when it came to playoffs, hence why i added in “DCMP/CMP’s in general” instead. I might poke at it later tonight if nobody else does, specifically for only DCMP/CMP quals.
2022 especially seems like an outlier here (and kind of 2017 St. Louis?), and might make sense that this subtle bias only manifests itself at the highest levels of play. It seems clear that it had no noticeable effect in quals. Unfortunately I didn’t get to see a field in person this year, so I can’t say more about that.
P.S. It would be nice if the yearly TBA insights could tell us what % of matches the blue alliance won in quals/playoffs/round robin, what alliance seeds are winning tournaments, etc.
If you use the binomial probability, there’s around a 1% chance that this happens on any einstein (assuming that its actually even) and there’s around an 11% chance that we’ll have an einstein that’s at least this weird after 7 tries.
15 matches is just a really small amount, e.g. if you flip a coin a bunch you’re eventually gonna have a streak that seems like it had to have been faked. There’s no way that we can take any probabilistic meaning from that.
Looking at the overall quals data doesn’t really help either. If the 254s of the world are evenly distributed across alliances, both alliances will probably win about the same amount. Unless this seeming blue advantage is very blue favored, the distribution of teams across alliances will have a much larger impact on the overall winrate of each alliance. There are just too many confounding factors here to really trust the stats data as the end all be all of this particular oddity.
Not totally convinced that the cable protector is such a big swing from the data presented, but I can’t think of any good reason not to have a dummy cable protector. We should strongly bias towards symmetry for both alliances.
