Thought I'd share a battle I just lost in Tark to vandydave (as A+ UNC to B+ Mizzou), both as a learning experience (it certainly was for me), and because I think it really crystallizes what I (and, I presume, some other coaches) really, really dislike about 3.0:

Recruit was Brandon Phillips, an SG out of IL. I had a preference advantage of +1 (he wanted to play, we both offered 15 min early, we both were very good on wants success, strong conference, and long term coach, but since he wanted to play far away from home, I was neutral, vandydave was bad). We both maxed out CV/HVs -- I spent 784 APs, vandydave spent 1200. So, apparently 1 of 2 things happened: (1) 416 APs outweighs a full letter grade of prestige and a +1 on preferences; or (2) it didn't, but the RNG came up vandydave's way. My problem is not with the first outcome -- although I would disagree with it, if WIS wants to program it that way, hey, it's their game that they're running into the ground. My issue is that we just don't know, because seble for some reason inserted randomization into the recruit's decision-making, so you can't actually learn from battles. It's just a giant shoulder-shrug. In my view, it takes a lot of the strategy out of recruiting.

I'd be interested to hear if anyone has had similar experiences -- or alternatively, if I am way undervaluing the import/effect of APs (which I may be).

For the record - I gladly sent John all of my recruiting efforts - it's nonsense that I won this battle - it's not everything wrong with recruiting but it certainly is a good example of why I want no part of new HD, keep in mind I won this "coin flip" and I still feel this way.

This is pricelesly the issue with the randomness element. Awful. Takes away skill.

It's both. You are substantially undervaluing the weight given to AP. They are outrageously overweighted and not scaled appropriately by Division or prestige. What was each school at the time of signing, both VH or VH & H?

The second part is, from the analysis provided in the Beta Forums of a few battles, even a VH v. H battle gave fairly significant odds to the H school. Admin was not terribly clear on how the cutoffs and odds worked, but I suspect that anything worse than 70/30 would put the lower team down to "moderate" and they would be removed from the RNG completely (even the the calculation without the thresholds would give them some chance). In other words, my sense is that a VH v. H spread would be about a 70/30 probability, which I view as very significant odds for the lower considered team.

The preferences clearly help unlock actions earlier and they act as some multiplier on effort. Given the lack of tourney cash and the fact that nothing was done with prestige, Prestige (both team and conference) is effectively nerfed...even though WIS can honestly state that "prestige was not changed". In the end, however, with two schools with that prestige and from BCS conferences, I would think admin would tell you that it is functioning as intended and, if they chose to be honest, that result was probably a virtual coin-flip (somewhere around 55/45 or so). So, you cannot make much out of it.

To me, the success or failure of the game depends on admin realizing that the overall structure of the recruiting game is ok, but the weight of all these factors and the resulting thresholds and probabilities are screwed up.

And, if we got rid of the randomness, then Vandy would have known he won the battle on skill, and utilized that strategy of tons of APs going forward when necessary. Now, he still doesn't know what worked or didn't work for him. I just audibly sigh.

but its so much more fun...

I agree w/ rogelio that AP are overweighted. based on battles in beta and the early goings in HD3, 400+ AP is a pretty substantial difference. with a limit placed on HV, if 50 AP = HV (wasn't that figure floated somewhere?), that's like 8 extra HV to compensate for the prestige/preference match, and that's significant. maybe enough to make it a toss up? it is frustrating not to know. being able to sort out and adjust the logic in these battle outcomes is really a major reason HD3 should have stayed in beta a little longer.

the whole "coin flip" thing really is a matter of opinion. i think it's valid to design a game where the person who had exactly 50.00001% of the effort should win every time, which is what a lot of people want. i think it's also valid to say the person who had 49.99999 chance should have a chance to win as well since this could increase the competitiveness in battles, which is something I've seen a lot of people say they like about HD3 as well. i mean, it's just as much 'skill' to ensure you have 50.1% of the effort as it would be to work the new system to get, say, a 40% chance at multiple recruits. it's all in how you look at it.

Posted by rogelio on 9/30/2016 10:52:00 AM (view original):
It's both. You are substantially undervaluing the weight given to AP. They are outrageously overweighted and not scaled appropriately by Division or prestige. What was each school at the time of signing, both VH or VH & H?

The second part is, from the analysis provided in the Beta Forums of a few battles, even a VH v. H battle gave fairly significant odds to the H school. Admin was not terribly clear on how the cutoffs and odds worked, but I suspect that anything worse than 70/30 would put the lower team down to "moderate" and they would be removed from the RNG completely (even the the calculation without the thresholds would give them some chance). In other words, my sense is that a VH v. H spread would be about a 70/30 probability, which I view as very significant odds for the lower considered team.

The preferences clearly help unlock actions earlier and they act as some multiplier on effort. Given the lack of tourney cash and the fact that nothing was done with prestige, Prestige (both team and conference) is effectively nerfed...even though WIS can honestly state that "prestige was not changed". In the end, however, with two schools with that prestige and from BCS conferences, I would think admin would tell you that it is functioning as intended and, if they chose to be honest, that result was probably a virtual coin-flip (somewhere around 55/45 or so). So, you cannot make much out of it.

To me, the success or failure of the game depends on admin realizing that the overall structure of the recruiting game is ok, but the weight of all these factors and the resulting thresholds and probabilities are screwed up.

Agree 100% with this post. OP is undervaluing weight of AP for sure I think. That is a big difference in AP used by Vandy so that was the main reason he won in my opinion. Or I should say it gave him the best odds of signing the player so I'm not surprised by the outcome based upon what we saw in beta.

I also agree that the 70/30 is probably pretty accurate and what I was thinking of for rule of thumb. Moderate you have 0% chance and at high you have a minimum 30% chance. No idea how accurate that is but it's a good start I suppose.

Finally, I also agree that they need to back off the weight of the factors and the signing probabilities. If there's a significant gap between effort like a VH vs H (whatever that is determined to be) then the 2nd place team should really only have like a 10% chance. It needs to be low. It needs to be rare for it to happen if that's the way they want signings to work or else it's just going to be a headache.

I do not object to uncertain results in battles and think some element of randomness is okay. This creates a different tactical and strategic environment.

Some may or may not like that.

What I find very frustrating is the greater amount of time and effort required to do what I think of as a "good" job of scouting.

And the failure to address EE is also frustrating.

add the major increase in effort to the frustrating features - and then the randomness adds frustration.

IIRC, the drop of the lesser team to moderate is closer to 60/40 than to 70/30. I think that seble explicitly said in the beta forum that a 65/35 battle would have the 35 team at moderate.

I don't have any problem with the randomness in signings. I actually think it is a positive addition to the game. I do think APs are massively overweighted right now. I'd love to see that as a question in the DevChat. Has anyone asked it yet?

A follow-up on my original post -- we were both very high on the recruit. And to be clear, I do not (necessarily) object to the idea of 416 APs outweighing the prestige and preference advantage. I just want to know -- or at least, have a better sense -- of whether that's the case. I think jpmills comment above is exactly correct -- there's really no strategy anymore; it's just a crapshoot because there are too many variables and unknowns.

FYI - this was Very High vs Very High

Posted by rogelio on 9/30/2016 10:52:00 AM (view original):
It's both. You are substantially undervaluing the weight given to AP. They are outrageously overweighted and not scaled appropriately by Division or prestige. What was each school at the time of signing, both VH or VH & H?

The second part is, from the analysis provided in the Beta Forums of a few battles, even a VH v. H battle gave fairly significant odds to the H school. Admin was not terribly clear on how the cutoffs and odds worked, but I suspect that anything worse than 70/30 would put the lower team down to "moderate" and they would be removed from the RNG completely (even the the calculation without the thresholds would give them some chance). In other words, my sense is that a VH v. H spread would be about a 70/30 probability, which I view as very significant odds for the lower considered team.

The preferences clearly help unlock actions earlier and they act as some multiplier on effort. Given the lack of tourney cash and the fact that nothing was done with prestige, Prestige (both team and conference) is effectively nerfed...even though WIS can honestly state that "prestige was not changed". In the end, however, with two schools with that prestige and from BCS conferences, I would think admin would tell you that it is functioning as intended and, if they chose to be honest, that result was probably a virtual coin-flip (somewhere around 55/45 or so). So, you cannot make much out of it.

To me, the success or failure of the game depends on admin realizing that the overall structure of the recruiting game is ok, but the weight of all these factors and the resulting thresholds and probabilities are screwed up.

Rogelio, are you saying that it is your understanding that if both teams are very high, it is a true 50/50 coinflip?

If prestige hasn't changed should the attention points been a wash. In 2.0, at DI each 1/3 of a letter grade in prestige was thought to be about a 15 - 20% multiplier (I believe) depending on where in the range of the 1/3 of the grade each team was. So John should have had about a 50% advantage depending where on the A+ and B+ range both teams were. That would make the attention points a wash. I assume that the HVs and CV of the A+ school have more value than the B+ giving John and advantage, then the +1 in preferences should have given him a bigger advantage. If that thought logic is correct (and I'm certainly not sure it is) then John had the advantage and just got a bad result.

I really wish after the battle you could click on the considering tab and see the actual percentage chance that each team had to win. They keep who was high and who was very high listed anyway, just put a percentage next to the team. I personally could live with "I was a 90% favorite to win the recruit and got a bad result but at least I know I did what I needed to do to usually win" much more than "I lost, but I really have no idea if I was ahead, behind, or how close it was".

Both Very High would not be true coinflip, but I would bet it would be within that 55/45 range. Sure.

Tarv's post, above, makes me even more confident of that range.