|
|
|
![]() |
|
|||||||
|
||||||||
![]() |
| Thread Tools | Rate Thread | Display Modes |
|
#31
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
From my own experience, the only qualifying match we lost was the one where our robot was not prepped before the match properly and our autonomous didn't work. The 10 point bonus is a huge difference is many cases, as it seemed most match scores were around 40 points. It also seemed many alliances could not score in autonomous, so even without getting many points you could get the ten point bonus. Except for 1610 and a few others, most of the high shooters were fairly inaccurate in autonomous, and the low dumpers worked well. Shooting low, we were making about 6 points in autonomous, and we then tweaked it to score 9-10 in autonomous. The only time we lost autonomous when it when it was working was against 1610.
|
|
#32
|
|||||
|
|||||
|
Re: A statistical analysis of the "autonomous advantadge"
Winning the autonomous mode was a huge factor at the BAE Granite State Regional. In matches where our alliance won autonomous, we never lost the match. We also never came back to win a match where we lost autonomous.
However, we did have two matches which ended in a tie. (In one of the ties, we had lost autonomous and in the other there was no autonomous winner.) It also became clear that a real key to winning the autonomous period would be using automatic camera-based targeting in order to be able to make high goal shots despite encountering defense in autonomous. After we finally got our automatic camera-based targeting working (not until the last qualification round match on Saturday morning) we didn't lose any more autonomous periods. |
|
#33
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
30 was a magic number of sorts. Judging by penalized scores... 30 pts would have won 54 of 59 (92%) matches on Friday at VCU; and would have won 73 of 86 (85%) both days. The highest losing scores were 46, 44, and 43. KA-108 ![]() |
|
#34
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Definately have to agree with everything Joel said. Its very difficult to say wether winning the match due autonomus mode is due to being a strong team, being on defense first or the 10 point bonus. I feel its a mix of all three, but there is definately a strong coelation. I kept track of stats of the semis at VCU and here is what I saw.
Matches in the semi both went to 3 while the finals went to 2. 5 out of 8 matches, the team that won auto also won the match. Of the other three, the winning auto team "lost" with one being a tie, one being a DQ, and one for the 10 point bonus on the ramp. About the game, scores were rather low compaired to past years. I did not compute an average, but it would likely be around 30 with 50 being the high end and 70 being amazing. Most teams seemed to ether miss all their shots (in auto as well), hit a few while being marked, or scoring many quickly. Often the announcer went: Score.. score.. score, score. I would like to say that was very very helpful cause the camera was not always in a good angle to see what was happening. edit: also, most teams missed the majority of their shots in automode. Teams take note... 1 point can often make all the difference wether it was 2 to 3 or 10 to 12! There was very little scoring in the low corner goals, outside of auto. At VCU at least, a strong focus was on shooting and "jocking for position." Several times the corners were not well guarded, and I think a quick dump would have proved better both in terms of score and time off the clock. Last edited by Martinez : 06-03-2006 at 13:09. |
|
#35
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
6 shots in autonomous is impressive. But did they do it consistently? Our team (319) sunk 7-10 balls EVERY autonomous. We only lost in the finals because the other alliance managed to knock our robot off course, so it couldn't find the light in time. |
|
#36
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
There is no doubt that our auto mode helped...When we shot in auto and scored it made it all that easier to have enough points to win a match.
But even when we did'nt score in auto other things we did could and did make up for this. There was one match (before finals) where we we're unable to score in auto and we're able to pick up balls fast enough to go back and score and still win the round. I'll let out a little secret here...we have multiple auto modes for our robot. Between who we were going against and what our alliance decided, decided what auto mode we used...Most of the times this worked...sometimes it did'nt...and some teams just never caught on to what were doing and were not able to counter this. Looking back..were there better robots with auto than ours...yes were there better shooters than ours....yes also.... but in the end it came down do our alliances and how we played the game. |
|
#37
|
|||||
|
|||||
|
Re: A statistical analysis of the "autonomous advantadge"
Ok, everyone. There's two days until Week 2 kicks off, so here's everything I've got so far:
An electronic (XLS) version of the spreadsheet A printable (HTML) version of the spreadsheet Take some of them to your Week 2 event on Friday and Saturday and start recording how matches go. It's OK if you don't get every single match, even a little bit of data will help build an overall information set we can use. If your team has some extra kids sitting around just watching, this could be the perfect job .After you fill it out, scan it or photograph it or type it in, and send it to aim.high.autonomous [at] gmail [dot] com. I'll compile a master list that I will post here, and then we can all analyze the heck out of it. |
|
#38
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
Ever since auto was introduced in 2003, teams have been writing dozens and dozens of auto routines that just never are needed in competition. Yet they sit at the ready on their robot. Who needed a routine last year to stop someone from scoring a vision tetra on the center goal? Since it didn't happen, all those programs sat idle, but I know a number of teams had one. You are forced to look at what options are available both offensively and defensively in autonomous, and create the appropriate responses to an opponents potential action. This usually means a high number of auto modes. This versatility also makes your bot an easier "fit" when selecting robots for playoffs. |
|
#39
|
|||||
|
|||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
I was at most 5 feet from the field for every match played at the NJ regional. And from my vantage point it was clear that most teams didn't figure out how to play this game until Saturday morning, and several never "got it" at all. This is a very complex game from a strategy standpoint and there are MANY valid strategies that can win. There is no question that there is a correlation between winning autonomous and winning the match. But as others have said, that could easily be due to the fact that the autonomous winner is simply the better alliance. Most of the capabilities needed to win auto are the same things that will ensure better performance thru-out the remainder of the match. Through the season as more teams learn "how to play the game" I predict autonomous wins will be LESS correlated to match wins. |
|
#40
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
|
|
#41
|
|||||
|
|||||
|
Re: A statistical analysis of the "autonomous advantadge"
Quote:
There are many different autonomous scenarios that call for different programs. Any autonomous programmer knows that. It wouldn't surprise me if Bharat and the gang from Team 25 have a few more Aces up their sleeves for Las Vegas (Hopefully, that will not get them kicked out of town ) |
|
#42
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
If you check out my statistics thread over here, you can see that the median winning margin was 11 points. The 10 points given for autonomous could very well be the difference between winning and losing for almost 50% of the matches.
|
|
#43
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
Maybe this is just me, but at the UTC regional this year, there did not seem to be as large, if any, of a correlation between winning autonomous and winnign the game.
|
|
#44
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
in the FLR that made a big impact and was one of the reaons we won the regional and the cooperation between teams and the defensive/offensive auto modes
also joel from 229 figured out that it was basically whoever went on defense first which was either who won auto mode or if no one scored any points then whoever the computer picked |
|
#45
|
||||
|
||||
|
Re: A statistical analysis of the "autonomous advantadge"
I believe that winning autonomous SHOULD be a huge factor in winning the match -- winning autonomous takes alot of mechanical and programming skill, and to have a window of 10 seconds, to be able to shoot balls accurately or defend automatically is vital to winning. After that, most of the game relies on great driving and good luck. In future years, I think that there should be even longer autonomous modes
|
![]() |
| Thread Tools | |
| Display Modes | Rate This Thread |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| STAMP (Statistical Team Analysis of Match Performance) is off and running! | Ethulin | Scouting | 6 | 04-02-2006 23:12 |
| Statistical Analysis of Regional Competion Scores | rourke | Regional Competitions | 9 | 08-04-2004 01:05 |