Following the MN Signature event our team could not help but notice the points awarded for the autonomous bonus is significantly higher, percentage wise, than past games. It so happened there were more upsets in the elimination round than I’ve ever witnessed from previous MOA competitions. The stats from eliminations revealed every team that won auto, regardless of rank, won the match except 1 (DQ’s were not included in the data).
In SU average scores ranged low to mid 100’s (early season) with a 10 point auto bonus and OU (late season) with scores in the 70-100 range with an auto bonus of 8. The average scores for HS ended up being between 25-35 and the 6 point auto accounts for 17-25% of total points scored. Historically, the percentage of auto during early seasons didn’t typically exceed 12%. It seems as though there is a significant disadvantage to the team not winning auto, more-so than in previous games, to where it was nearly impossible for teams during eliminations to overcome the deficit.
Do you think the GDC should reevaluate the points for auto bonus or do you think the point value is where it should be?
This is a great point. Ive been looking at the balencing for high stakes a lot and there are a lot of interesting descions like this. For example a tier 3 climb and the high stake used to be 15 points (one less than two stakes) and is now 17 points (one more than two stakes). So I do think it is intentional that the AB is worth the same two top rings or a tier 2 climb.
In terms of the autonomous bounus determining the results, I think this is more due to how far ahead a good auton puts you. You could score 17-23 points with a good auton and matches avaraged around 30 points, so it is quite possible that most of the reason the bounus was a good indicator of final results is because, without many descore mechs or climbs past tier 1, a team could very easily win just off of what they did in auton and a little corner defence. Also, in high stakes autons can touch 1/4 of the scoring objects, compared to over under where autons could at most affect 1/10. So it makes sense that autons are much more important this year.
On the other hand, the bonus does come more into play when the autons are close in score and one team is 6 points up even though there was only a 1-3 auton point difference and the match is won by 1-5 points, but this was always the point of the AB, to make the autonomous more than just a challenge 15 second head start, but somthing that impacts the results of the match.
In a slight improvement over last year, the Tournament Manager overlay now shows the point-breakdown of a match, including who won auton. One could rewatch the MoA (sorry, MN Sig Event) livestream and sus that out; unfortunately the data available from RobotEvents does not provide that level of detail.
I believe I saw someone did an analysis of Autons in Elims at MoA and found a 94% correlation.
Based on my observations at Mall, I would agree that 6 points for Auton in this game is far too high.
The median winning score is 30 and the median losing score is about 15, and the median margin of victory is between 11 and 12, meaning half of matches could have been swung by auton.
That said, I think this game has significant game-play issues above-and-beyond the auton score.
Teams should have outfitted their robots with tents and brought marshmellows for the amount of corner camping seen. I think that the risk-profile of wall-stakes is terrible. They are easy to defend, and have the risk of accumulating SG4 (Ring-Falling-Outside-The-Field) violations leading to a DQ. The amount of time (both in design/building and on-field-execution) of the climb, for the amount of points available, is also poorly balanced, IMO. As it stands, matches really only consist of 15 second autons, 10 seconds at the start of driver and 10 seconds at the end. The middle 85 seconds of a match are incredibly dull. I can’t think of a less interesting VRC game to spectate.
I anticipate/hope for major scoring changes in the 9/3 update.
This is the real problem. Even a 4-point auton bonus could help level things out.
Yep. No doubt about that. I was going to start a Match Strategy Evolution thread like the one we had last year, but there isn’t much match strategy to discuss. I’m very disappointed with the quality of this game.
However, there is hope. Robots will eventually get good enough to the point that it is more lucrative for them to play offensive rather than camp the corners. This is when there will be some match strategy to discuss.
I think it is a big win - focus on coding!!! but seriously, early season and mechanical capabilities are not fully developed… I look forward to seeing this season progress!
I agree with all of you, but just a reminder: correlation is NOT causation.
The auton and win correlation could be because of a third factor, such as the fact that teams made a good robot or practiced much also had a good auton.
I think one event is too few to assess, particularly as teams did not come prepared to score by all methods. (Awesome builds, but its August.)
Very few attempted wall stakes (SG4 is rough, but…) – up to +20 points
No level 2 or 3 climbs – up to +18
No high stakes – +7
That is 45 points, and makes the 6 point auton bonus come down to the 10% range.
Most importantly, teams now know they need a strategy to remove opponents from a corner. If you have that tool in your kit, you can fill your possessed goal then do other tasks. If your opponent wants to camp, the field is yours to command.
Sometimes the GDC intends for a game to be autonomous heavy as a part of the gameplay.
As for the other components and issues, like every season there will be things that need to be fixed early season and then the GDC will patch and reinforce the weakest links of the game manual.
I wonder if protecting wall-stakes, similar to the hang, would alter the risk-reward enough for wall-stakes to matter. Something like “Robots attempting to score on wall-stakes are protected while they, or a Ring in their possession, is contacting a Wall Stake”
There’s still the inherent risk of failing to get the Ring on and having it fall out the field, but at least that wouldn’t be induced by an opponent ramming into them.
With Negative Zones, I’m still not sure that scores will reliably get into the 40s. Especially with climb rules seemingly ruling out the quick-elevation mechs from last year. My guess is that a Level 3 hang would probably take 15-20 seconds to execute under current rules. Again, I could be proven wrong
I think that protecting the wall stake play can be fixed by modifying SG4.
I also cannot see anyone bothering to high climb if negative zones remain unprotected. “Do I climb for 12 points or make a 16 point swing in the score?” is not a hard question to answer. By the nature of the game, there will be mobile goals loose during the last 10 seconds, and placing them in the negative corner will be more points than climbing.
What is even the point of the ladder if negative zones are unprotected.
Let robots climb with MoGos (this may have other undesired knock-on effects)
Bump the value of the climb tiers (maybe 6/12/18)
High Stake needs more value
Protecting all the zones for the Mogos will leave 10 seconds (or whatever potential change to the endgame time limit there may be) where teams will almost be forced to do something other than Mogos (well, they could continue to do nothing).
I would like to see the buddy-climb rule from VEXU come to V5RC HS, though climb in VEXU seems OP.
One way to fix this issue would be to make the positive and negative corners only apply to the top ring of the stake
This would incentivize going for the top ring and make it not a near automatic loss if you lose control of both positive corners while also lessening the reward for camping.
I reckon they do need to change 6 point to 4 point auton bonus. Because I feel that while auton should be important, Even with 4 points it is way more significant than all previous games. and the 1:45 driver period is still were teams should be winning or losing, and right now, there is too much correlation between auton win and game win. Autonomous should just be a small boost, not a massive blow to the other team. This change I feel would make games more fair and exciting.
