Championship Event - Where the "Random" Match Sorting Really "Shines"

[Jacob Plicque]

Originally Posted by Bongle

One easy solution to the tiering problem: Rather than split the teams into 3 tiers based on a sorted order, just randomize your input list so that the three groups will be approximately equal in quality. You would still have the problem of never playing with anyone in your tier as an ally, but at least you wouldn't have the big helping of rookies and hurting of veterans.

So rather than

1 4 7
2 5 8
3 6 9

You might have lists like

1 6 5
3 9 4
7 2 8

On average, they will even out to have the same number of powerhouse teams.

The whole tiering thing seems to be a wierd solution to a problem that didn't really exist anyway. Now instead of a few teams sometimes legitametely complaining that they got a bad deal in the pairings, you've got a whole group (the veteran teams) who get a bad deal, and they can prove it.

I suggest a modification to your idea which sorts the teams the same way as the divisions at the championships.
Current method
V M R
1 504 1207
2 505 1208
3 506 1209
A V team can never play another V team in the random pairings
Division Method
A B C
1 2 3
4 5 6 Low Pool
7 8 9
501 502 503
504 505 506 Mid Pool
507 508 509
1201 1202 1203
1204 1205 1206 High Pool
1207 1208 1209
This distributes the allegedly strong lower number teams (< 400) into three groups. Now a veteran team will still be paired with a midrank and a rookie in its division but it has the possibility of playing against 67% of the veteran teams

[Faith]

Originally Posted by GaryV1188

The problem is deciding which teams get into which tiers. Team number is not the way to go. If you go to a regional with few rookies, the 3rd tier gets distorted - such as including 1114 at GLR. At the other end, there are teams with very low numbers who have a poor year, or who have had a long-term decline due to various factors (losing sponsors or mentors comes to mind) and are just shells of their former glory.

My team, 1100, was actually in the lowest tier at Boston. At least I think we were, since the lowest number team we played with was 1403, and we only played with teams under 1900 twice (once was 1403).

[The Lucas]

Originally Posted by Plooshiska

My team, 1100, was actually in the lowest tier at Boston. At least I think we were, since the lowest number team we played with was 1403, and we only played with teams under 1900 twice (once was 1403).

Indeed, you were the last team in the A (or Vet) pool. The lowest number you could have been paired with was 1350. That is not a fun position to be in but you faired well at 5-4 and got picked 9th overall. Congrats.

There is good news and bad news regarding Championships.

The good news is you will be around the middle of the B (or Mid) pool in Galileo.

The bad news is your first match is likely to be:

250 1100 2026 v 279 1126 2056

Ouch, Tech Fusion, SPARX, and Patriotics in one match. You have FLR champ 250 on your side, but that is a tough task for even their defensive prowess. This the fake match 11 Alex Cormier starts talking about starting with this post

[JB987]

We'll take the preceeding scenarios over what we had in Vegas...team 987 was never allied with a team number less that 1000 during the qualification rounds...

[AcesPease]

The championships are over, and I thought it might make more sense to comment about the match schedule on an existing thread, rather than a generic championships likes/dislikes thread.

We were on Curie and it appears that the schedule achieved what it set out to do. A few observations:

1) Old and new teams were very mixed in the final standings and the alliances that were formed were all very close to each other. A little better strategy or luck and any of the alliances formed on Curie could have won that field. Watching an 8th seeded alliance win it all on Einstein, I have to assume the situation was similar on the other fields.

2) There were 8 rookies in the Championship alliances. For the record we teamed with a rookie at all three of our events this year. I don't know, but I doubt there have ever been this many rookies in the eliminations since the four division system started.

3) This game was very much a game of alliances. No machine was able to dominate unless their partners could help in some fashion. We won many rounds last year almost single handedly. We did not win any rounds this year unless we had help from our alliance partners, I suspect most people had similar experiences.

4) The match schedule employed seems to increase the likelihood of having a tough time in the qualification rounds. At one regional we almost breezed through quals because of excellent partners. At the next regional we struggled to win half our qual matches and we were almost left out of the elmins despite some excellent scoring by our team. At the championships we had a lot of close matches and won with a lot of good play by our partners. According to our scouting there were half a dozen teams on Curie ranked below 50 that belonged on the short list for picking. I know there have always been a couple really good picks way down in the standings, but almost a third of the top 24 sounds like a lot.

5) For the record I don't like the algorithm because in a field of 86 teams we should not have to face the same team twice. We were against one low numbered team twice, we were with a mid numbered team twice, and we were against and then with four more teams. That means we saw 29 different teams out of 86 when we could have seen 35, if the algorithm focused more on mixing up the match pairings.

FIRST seems to have accomplished what they wanted with the algorithm (see 1 and 2). But by limiting the variety of partners too many teams get locked into unfortunate schedules.

[Jeremiah Johnson]

I saw a lot of teams get bitten by the "random" algorithm monster again. This hurt a lot of great teams dearly... some super-great teams were waaay overlooked. Some teams weren't even given a chance to impress because they had partners that couldn't compliment each other... but that's how the game's played.

For the record, team 648 played against team 16 three times. Of 86 teams, what are the chances? I don't know the stats like that for other teams, but I'm going to assume we weren't the only ones that that happened to.

[dlavery]

Originally Posted by AcesPease

FIRST seems to have accomplished what they wanted with the algorithm (see 1 and 2). But by limiting the variety of partners too many teams get locked into unfortunate schedules.

Again, just to set the record straight, the match pairing algorithm is NOT what FIRST wanted. It was implemented by the software developer without the concurrence of FIRST. It was not corrected during the season, despite multiple FIRST requests.

FIRST knows the match pairing algorithm is a problem. They know it is an intolerable situation that caused significant problems with allowing teams to fairly display their capabilities and performance. It will NOT be continued next year.

-dave

[sporno]

Originally Posted by Budda648

I saw a lot of teams get bitten by the "random" algorithm monster again. This hurt a lot of great teams dearly... some super-great teams were waaay overlooked. Some teams weren't even given a chance to impress because they had partners that couldn't compliment each other... but that's how the game's played.

For the record, team 648 played against team 16 three times. Of 86 teams, what are the chances? I don't know the stats like that for other teams, but I'm going to assume we weren't the only ones that that happened to.

yes 540 got bitten really hard at championship and regionals when they had to go against 539 every single match.

[IndySam]

Originally Posted by dlavery

Again, just to set the record straight, the match pairing algorithm is NOT what FIRST wanted. It was implemented by the software developer without the concurrence of FIRST. It was not corrected during the season, despite multiple FIRST requests.

FIRST knows the match pairing algorithm is a problem. They know it is an intolerable situation that caused significant problems with allowing teams to fairly display their capabilities and performance. It will NOT be continued next year.

-dave

I hear what you’re saying Dave but it was a problem from the first day of the first regional and it was never fixed and it should have been.

You can blame the developer all you want but the buck has to stop with FIRST.

[dlavery]

Originally Posted by IndySam

You can blame the developer all you want but the buck has to stop with FIRST.

Perhaps. But I have to give them a little bit of slack when they tell the developer to make a change to the system (for which they are paying a significant price), and the developer just says "no - I won't do it. My way is so much better, and you (the customer) are just not smart enough to understand that."




.

[Andy Baker]

Originally Posted by dlavery

Perhaps. But I have to give them a little bit of slack when they tell the developer to make a change to the system (for which they are paying a significant price), and the developer just says "no - I won't do it. My way is so much better, and you (the customer) are just not smart enough to understand that."

I have an idea: We all can send this developer a note, tell him how much we liked or did not like match generating program he wrote.

Is it appropriate to post his email address?

Andy B.

[Richard Wallace]

Originally Posted by IndySam

...
You can blame the developer all you want but the buck has to stop with FIRST.

Originally Posted by dlavery

Perhaps. But I have to give them a little bit of slack when they tell the developer to make a change to the system (for which they are paying a significant price), and the developer just says "no - I won't do it. My way is so much better, and you (the customer) are just not smart enough to understand that."

Thank you, Dave, for getting this situation out into the open for all of us to see. I have understood for a couple of months now that the contractor FIRST hired has his own agenda. I think we've all been waiting for the real FIRST (the board and the GDC) to raise its voice about that.

My primary concern about the algorithm, which I expressed to my own favorite board member just after Week 1, was never that it is unfair, capricious, stupid, flawed, bogus, etc; although it is all that. My concerns were that it did not reflect the published intent of FIRST as stated in the manual, and that it in fact reflects the agenda of a software system contractor who fails to understand that role; i.e., who openly defies direction from his employer. The $20B firm that employs me would not employ such a contractor more than once.

[Nuttyman54]

Originally Posted by Andy Baker

I have an idea: We all can send this developer a note, tell him how much we liked or did not like match generating program he wrote.

Is it appropriate to post his email address?

Andy B.

As much as I like this idea, I do not think posting an email address is the right thing. This person, while they may have not listened to FIRST, does not deserve the massive amounts of hate mail that will invariably be sent. Not to mention, posting an email address on a public forum is generally asking for trouble.

I do, however think that this particular software developer does need to get feedback from the teams. I cannot think of an easy way to do this currently, but I'm sure someone out there can.

[Dave Flowerday]

Originally Posted by dlavery

Perhaps. But I have to give them a little bit of slack when they tell the developer to make a change to the system (for which they are paying a significant price), and the developer just says "no - I won't do it. My way is so much better, and you (the customer) are just not smart enough to understand that.".

If this is the case, then I guess you're right that FIRST should get some slack. However, isn't this the same (or similar) story that we've heard a few times in the past, such as with Hatch? At some point FIRST needs to accept the fact that they are putting themselves in this position by selecting these vendors in the first place.

[meaubry]

Andy,
I'm sorry, but posting the persons email address is NOT appropriate (at least not on this website). If we allowed it, just imagine how many others would need to be posted whenever someone had a beef against something. We really don't want to be associated with that kind of behavior.

As much as I disliked the algorithem, I think Dave has done a great job trying to explain that the supplier failed to, and refused to meet the customers (FIRST) requirements.

As in business, that supplier should not be used again, if what David has stated is correct, and I have no reason NOT to believe him.
If I ran FIRST, I'd not ask him to return and I'd throw out the algorith never using it again.

We all understand where the "buck stops", I'm sure that they learned something themselves this time and we will not see a repeat next year.

Thanks Dave for trying to clear up what happened. I appreciate knowing this.

Mike Aubry