(Seemingly) Irrational alliance selection decisions?

[jajabinx124]

I agree with most of the responses to the OP's question.

Most of the irrational/illogical decisions I've seen come from a team that probably didn't scout as well or pick based on bias/experience working with the team.

Scouting and Strategy is a hard system to run for teams during competition and it's harder to manage than imagined. Many teams have different systems, organization, different amounts of manpower, resources, and etc. that probably affect how well their picking is.

I'm not sure if I've seen any declining based on this reason, but I declined the 2nd seeded alliance's invite (as 2052's alliance selection representative) at the 2015 10,000 lakes regional because I wanted our team to captain the alliance to get to the finals to guarantee ourselves a wildcard. I had a few other reasons to decline because I knew I had 1/2 better options to pick after declining them, but getting the wildcard was the pushing factor since the first wildcard is awarded starting with the opposing alliance captain to the wildcard team on the winning alliance.

At the 2015 10,000 lakes regional, the 2 best teams joined together to form the no.1 seed (525 and 2502) and since 525 had won the Northern Light's regional previously, they had a wildcard. Strategically I chose to decline the 2nd seeds invite because I felt like I could built an alliance strong enough and go to the finals to meet the 1st seed and guarantee my team a spot at champs via wildcard. (which at that point it was pretty evident the 1st seed with 525 and 2502 was the powerhouse alliance that had the best shot at winning the regional)

Reason I decided to mention this is because after I declined on stage I'm pretty sure half of the people in the stands (including most of my own team members) were pretty confused as to what the heck I as doing and probably thought I was being irrational.. but I do think this is more of a unique reason that a team chose to decline another team's invite.

[Rangel(kf7fdb)]

I think it's really hard to define what the "best" picks are. Especially since the goals for every team that picks is pretty different. One team might pick overall solid robots and just try to get as far as they can in eliminations whereas others might pick robots that are flaky but have the potential to take down whoever the strongest alliance is. Both of these approaches can backfire with either not having enough firepower or having bad matches early and getting knocked out in the quarters. Especially in games like Stronghold too, it may not be wise to pick the best scoring robot if there isn't enough synergy between the robots. For example, having 3 offensive robots scoring in the center of the outerworks is not ideal and are likely to get in the way. Conversely, the coordination might be good enough that this isn't even a problem. Overall, I think it's hard to really judge how strong picks actually are until the matches are actually played.

[TheBoulderite]

Everyone thought it was irrational at Central Valley 2014 when 254 picked 973, two places away from last place in qualifications. It really depends on what strategy you think is best.

[AdamHeard]

Originally Posted by TheBoulderite

Everyone thought it was irrational at Central Valley 2014 when 254 picked 973, two places away from last place in qualifications. It really depends on what strategy you think is best.

#SandBagFirstPick

Just glad we finally got to #SandBagFirstSeed this year for once...

[SoftwareBug2.0]

Originally Posted by jajabinx124

I'm not sure if I've seen any declining based on this reason, but I declined the 2nd seeded alliance's invite (as 2052's alliance selection representative) at the 2015 10,000 lakes regional because I wanted our team to captain the alliance to get to the finals to guarantee ourselves a wildcard. I had a few other reasons to decline because I knew I had 1/2 better options to pick after declining them, but getting the wildcard was the pushing factor since the first wildcard is awarded starting with the opposing alliance captain to the wildcard team on the winning alliance.

At the 2015 10,000 lakes regional, the 2 best teams joined together to form the no.1 seed (525 and 2502) and since 525 had won the Northern Light's regional previously, they had a wildcard. Strategically I chose to decline the 2nd seeds invite because I felt like I could built an alliance strong enough and go to the finals to meet the 1st seed and guarantee my team a spot at champs via wildcard. (which at that point it was pretty evident the 1st seed with 525 and 2502 was the powerhouse alliance that had the best shot at winning the regional)

Reason I decided to mention this is because after I declined on stage I'm pretty sure half of the people in the stands (including most of my own team members) were pretty confused as to what the heck I as doing and probably thought I was being irrational.. but I do think this is more of a unique reason that a team chose to decline another team's invite.

I also took a whack at reviewing alliance selection results from this year. The way I did it was to say that a pick made sense if there was some way in which to rank teams in which the one that they chose was in the top five of those still available. I found a ton of picks that I just could not account for.

I thought I'd see a couple alliances per event that was just doing pure OPR based picking or pure rank, but it turned out that for the events that I was looking at those weren't the case at all. And a lot of them didn't make any more sense after looking at scouting data. But I did find that there was a positive correlation between alliances that made picks that I could figure out and advancing further in the tournament.

It seemed that many captains picked so badly that they would have increased their odds of winning the tournament if they had just been picking based on rank. I can only conclude that are large number of picks are often made for reasons other than maximizing the odds of winning the tournament.

And I think there are sometimes good reasons for trying to do something other maximize odds of winning, like the one that jajabinx124 cited with the wildcard system. And similarly to the wildcard system, in the district system how far you advance in the tournament matters, not just winning or losing the event.

I'm not sure how common this sort of strategy is however. For example, during the PNW championship I knew exactly how many points my team had to get to give us a 100% chance of making worlds, but talking with some other teams in attendance it seemed like a lot of folks didn't really grasp the point system well enough to figure out how well they had to do to get invited to worlds, let alone what that would imply strategically.

One extra note I will give is that it was easier to predict selections as the level of play increased. For example, while some in the audience and even the MC found the alliance selection on Hopper surprising it really shouldn't have been a surprise to any of those involved.

[TDav540]

Originally Posted by SoftwareBug2.0

I also took a whack at reviewing alliance selection results from this year. The way I did it was to say that a pick made sense if there was some way in which to rank teams in which the one that they chose was in the top five of those still available. I found a ton of picks that I just could not account for.

I thought I'd see a couple alliances per event that was just doing pure OPR based picking or pure rank, but it turned out that for the events that I was looking at those weren't the case at all. And a lot of them didn't make any more sense after looking at scouting data. But I did find that there was a positive correlation between alliances that made picks that I could figure out and advancing further in the tournament.

It seemed that many captains picked so badly that they would have increased their odds of winning the tournament if they had just been picking based on rank. I can only conclude that are large number of picks are often made for reasons other than maximizing the odds of winning the tournament.

And I think there are sometimes good reasons for trying to do something other maximize odds of winning, like the one that jajabinx124 cited with the wildcard system. And similarly to the wildcard system, in the district system how far you advance in the tournament matters, not just winning or losing the event.

I'm not sure how common this sort of strategy is however. For example, during the PNW championship I knew exactly how many points my team had to get to give us a 100% chance of making worlds, but talking with some other teams in attendance it seemed like a lot of folks didn't really grasp the point system well enough to figure out how well they had to do to get invited to worlds, let alone what that would imply strategically.

One extra note I will give is that it was easier to predict selections as the level of play increased. For example, while some in the audience and even the MC found the alliance selection on Hopper surprising it really shouldn't have been a surprise to any of those involved.

It all depends on your goals. Some teams want to make semis, some finals, some win. Some just want to make the playoffs and have a good time.

Additionally, although it may seem to us (especially in hindsight) that a team's picks didn't make sense, they may have a completely valid reason. Some teams have untapped potential (I've worked with a few of those cases), others have (or at least think they have) good chemistry, others still have complimentary strategies. It may be an incorrect assumption or decision, but it's what that team believed at the time.

How often do teams make the right pick? There are enough variables to consider that really teams should be just looking to make their best possible pick. At the end of the day, doesn't the "right" pick depend on the result?

[Christopher149]

Originally Posted by TDav540

It all depends on your goals. Some teams want to make semis, some finals, some win. Some just want to make the playoffs and have a good time.

Some are at district champs and want to help certain teams get to CMP.

[Jon Stratis]

I wouldn't say any alliance selection decisions are irrational. Some are unprepared, some are based on faulty data, and some are based on different value propositions.

Perhaps the biggest thing to remember during alliance selection is that no one has perfect information. Every team representative down there is dealing with a state of asymmetrical information, in which every team is basing their decisions on different information. Sure, you can say that they all have access to the same information - match results, for example - but the process of collecting data and processing it into a usable pick list causes the data itself to undergo a different transformation for every team, and end up being viewed differently by every alliance representative. As a result, we end up with picks that may or may not agree with any given set of data or viewpoint.

Honestly, I find the uncertainty going into alliance selection rather exciting.

[TheBoulderite]

Another seemingly irrational alliance selection this year was when 1425 selected 1538 on Hopper. Don't get me wrong, 1538 is a good team, but they could have gone with a team like 971, 1323, 4334, or 4587. Anyone from 1425 want to chime in and tell me why you made the choice to go with the Cows? I'm not saying it's the wrong choice, I'm just curious.

[Citrus Dad]

Originally Posted by TheBoulderite

Another seemingly irrational alliance selection this year was when 1425 selected 1538 on Hopper. Don't get me wrong, 1538 is a good team, but they could have gone with a team like 971, 1323, 4334, or 4587. Anyone from 1425 want to chime in and tell me why you made the choice to go with the Cows? I'm not saying it's the wrong choice, I'm just curious.

As mentioned, 1425 was declined by 971. 1538 had shown flashes of brilliance, so to have a good chance at advancing 1425 need 1538 to become consistent at that level. Unfortunately for them, it didn't happen. This year the rest of the draft pretty much went to form in our division with no real surprises.

We were the beneficiaries in 2013, 2014 and 2015 of seemingly irrational picking at champs. We had gamed out the scenarios in 2013, and even anticipated the declinations. In the end, the other teams should have looked at where us and 1983 (#2 seed), and then a rookie in #3 and realized that the field was going to get scorched no matter what. In 2014, we had a heated debate about whether to pick 1114 or 971, and then no one picked 971. We were shocked. In both years we were very fortunate to have 872 and 1641 available for 2nd picks.

In 2015, it looked like we got 1671 due to oversight, but in fact the other alliances were making selections to match their overall alliance. Maybe they should have rethought their strategies, but what they did was rational.

In general our 2nd pick comes from between 10th and 16th on our pick list of 24, but we've gotten as high at 6th a couple of times and that's when we go "huh"?

We've seen some real problems at regionals, so I watch for inexperienced teams that might end up at alliance captains and help them with their draft lists if they need it. As SVR this year, I told 5700 that they were going to be an alliance captain--they had no clue. We found them another another more experienced team to develop their draft list. I don't want to see a team looking at the ranking board trying to figure out their next pick.

[youngace89]

One factor which has not yet been discussed but played a major role in our decision making at the 10000 Lakes regional this year is the role of seeding and the bracket.

We were pretty sure going into the regional that we didn't have much of a chance of winning against the great palindrome of 2052, 525, and 2502, so, as usual, our goal was to make it to the finals and get a wildcard, like we had done the 2 previous years,

We ended up ranking 5th. Another captain was selected ahead of us, so we moved up to the 4th spot, and then 2846, who moved up to third, selected us. Had our goal been to win the regional, we probably would have declined them and picked 2502 or 2823, who were better high goal shooters. However, we knew that if we declined them, we would end up as the captains of the 4th alliance, and we would have to play the dominant 1st alliance with 2052 and 525 before we reached the finals.

So we accepted and ended up making it to the finals and getting a wildcard.
2502 and 2823 paired up to form an alliance which had a much better chance of beating 2052 and 525 than ours did, but they had to play them in the semifinals because of their seed in the bracket, so I would still rather have been in our position.

[IKE]

Originally Posted by Citrus Dad

...snip...

In general our 2nd pick comes from between 10th and 16th on our pick list of 24, but we've gotten as high at 6th a couple of times and that's when we go "huh"?
...snip...

Do you typically have a second list for 2nd pick? IE are you looking for different attributes that might elevate tem XYZ from say 20th as a first round to say 10-16th as a second round?

Als, are you talking Worlds or regionals.

Ususally the highest off my pick list that does not get picked up is around #20 or so.

[Koko Ed]

The craziest situation I ever witnessed for alliance selection was at the 2009 FLR where 610 was a 13th seed and turned down an invitation from 1765 who had their best weekend in their history that weekend. 610 was taking a huge gamble that the other alliances were going to pick one another and they'd settle in as an 8th seed but for that to happen they would have to hope that 340 would pick their little sister team 424 which they didn't and thus 610's day was done.

[OccamzRazor]

Hey did anyone watch Curie division picks? 100% upset brackets this year in 2016. Every alliance selection made me cringe until the 7th or 8th seed. (which hilariously enough were division finals)

[TheBoulderite]

Originally Posted by OccamzRazor

Hey did anyone watch Curie division picks? 100% upset brackets this year in 2016. Every alliance selection made me cringe until the 7th or 8th seed. (which hilariously enough were division finals)

Curie was an anomaly this year. Some of the better-known teams like 3310, 1983, and 2848 weren't as strong at Curie as I thought they would be. However, they were good enough to be selected for eliminations, and the captains thought they could win with them. I didn't cringe, I was curious to hear the strategy behind making those selections.