Maybe it was a typo and they meant 30%. I just looked at the top 100 signings in Knight from the 1st session and then pulled out any of the signings that involved a VH and H. Here are the results.

#1 Zachary Edwards. 2 VH vs 1 H. VH won
#4 Dennis Davies. VH Vs 1 H. VH won
#6 Brian Stutler. 2 VH vs 1 H. High won
#8 Clinton Lent. VH vs H. VH won
#16 Curtis Thomas. VH vs H. VH won
#17 Gregory Morton. VH vs H. VH won
#21 Kyle Northup. VH vs H (the High is actually Seble himself!) . High won
#30 Richard Dunn. 2 VH vs H. High won
#37 Eugen Hillhouse. VH vs H. VH won
#52 James Hunnicutt. VH vs H. VH won
#55 Alfred Clarke. VH vs H. High won
#56 Emmett Lefebre. VH vs H. VH won
#61 Paul Woods. 2 VH vs H. High won
#65 Leon Clayton. VH vs H. VH won
#81 Timmy Felton. VH vs H. VH won


In summary, there were 15 battles. 10 of the battles went to VH and 5 went to High. What I find even more incredible is that THREE of High signings actually beat out TWO VHs.

So yeah, definitely not 3%.

Re-posted from another thread:

I know at first glance it looks like they're saying highs have a 3% win rate against very highs. But that's not what they said. Go back another sentence for full context:

"Only a couple worlds have begun to recruit but thus far a majority of the signed recruits have gone with the team that had the most interest as expected. Only 3% have gone with a High interest team over a Very High interest." (emphasis mine, obviously)

They aren't saying highs have a 3% win percentage over very highs. I suspect that's actually between 20 and 30%, and I don't expect them to disclose that figure, nor the target figure they have in mind; they probably would have done it already if they intended to. They're saying of the set of signed recruits, a majority go with the team with most interest (% not disclosed) and a subset of only 3% includes recruits that chose high over very high. It's not a win rate, it's a percentage of the overall signees that chose high over very high. The point being, while those upsets understandably get a lot of attention, they represent a very small number of overall recruits, because most signings don't involve a battle between high and very high.

I think their post is intentionally misleading. Like they are trying to make it seem like there is less randomness than their really it.

Posted by pkoopman on 10/1/2016 9:06:00 AM (view original):
Re-posted from another thread:

I know at first glance it looks like they're saying highs have a 3% win rate against very highs. But that's not what they said. Go back another sentence for full context:

"Only a couple worlds have begun to recruit but thus far a majority of the signed recruits have gone with the team that had the most interest as expected. Only 3% have gone with a High interest team over a Very High interest." (emphasis mine, obviously)

They aren't saying highs have a 3% win percentage over very highs. I suspect that's actually between 20 and 30%, and I don't expect them to disclose that figure, nor the target figure they have in mind; they probably would have done it already if they intended to. They're saying of the set of signed recruits, a majority go with the team with most interest (% not disclosed) and a subset of only 3% includes recruits that chose high over very high. It's not a win rate, it's a percentage of the overall signees that chose high over very high. The point being, while those upsets understandably get a lot of attention, they represent a very small number of overall recruits, because most signings don't involve a battle between high and very high.

Ah I see. Poorly worded response.

Well wtf, that's not what we wanted to know. Well, there were 66 signings of the top 100. And 5 went to a high team. So thats about 7%. So I suppose it's close.

But regardless. I guess we can still take the data from the top 100 signings of H over VH. With a small sample size, it's about 30%. But I find the fact that 3 of them were with just 1 high beating 2 VHs. That makes NO sense for it to happen that much.

Posted by Benis on 10/1/2016 9:14:00 AM (view original):

Posted by pkoopman on 10/1/2016 9:06:00 AM (view original):
Re-posted from another thread:

I know at first glance it looks like they're saying highs have a 3% win rate against very highs. But that's not what they said. Go back another sentence for full context:

"Only a couple worlds have begun to recruit but thus far a majority of the signed recruits have gone with the team that had the most interest as expected. Only 3% have gone with a High interest team over a Very High interest." (emphasis mine, obviously)

They aren't saying highs have a 3% win percentage over very highs. I suspect that's actually between 20 and 30%, and I don't expect them to disclose that figure, nor the target figure they have in mind; they probably would have done it already if they intended to. They're saying of the set of signed recruits, a majority go with the team with most interest (% not disclosed) and a subset of only 3% includes recruits that chose high over very high. It's not a win rate, it's a percentage of the overall signees that chose high over very high. The point being, while those upsets understandably get a lot of attention, they represent a very small number of overall recruits, because most signings don't involve a battle between high and very high.

Ah I see. Poorly worded response.

Well wtf, that's not what we wanted to know. Well, there were 66 signings of the top 100. And 5 went to a high team. So thats about 7%. So I suppose it's close.

But regardless. I guess we can still take the data from the top 100 signings of H over VH. With a small sample size, it's about 30%. But I find the fact that 3 of them were with just 1 high beating 2 VHs. That makes NO sense for it to happen that much.

Keep in mind that the considering tab reflects relative interest. When there are 3 battlers, it could be 35-35-30. We're going to see this more in the top 100 than anywhere else, because people aren't likely going to want to take those kinds of long odds farther down the line.

Posted by pkoopman on 10/1/2016 9:19:00 AM (view original):

Posted by Benis on 10/1/2016 9:14:00 AM (view original):

Posted by pkoopman on 10/1/2016 9:06:00 AM (view original):
Re-posted from another thread:

I know at first glance it looks like they're saying highs have a 3% win rate against very highs. But that's not what they said. Go back another sentence for full context:

"Only a couple worlds have begun to recruit but thus far a majority of the signed recruits have gone with the team that had the most interest as expected. Only 3% have gone with a High interest team over a Very High interest." (emphasis mine, obviously)

They aren't saying highs have a 3% win percentage over very highs. I suspect that's actually between 20 and 30%, and I don't expect them to disclose that figure, nor the target figure they have in mind; they probably would have done it already if they intended to. They're saying of the set of signed recruits, a majority go with the team with most interest (% not disclosed) and a subset of only 3% includes recruits that chose high over very high. It's not a win rate, it's a percentage of the overall signees that chose high over very high. The point being, while those upsets understandably get a lot of attention, they represent a very small number of overall recruits, because most signings don't involve a battle between high and very high.

Ah I see. Poorly worded response.

Well wtf, that's not what we wanted to know. Well, there were 66 signings of the top 100. And 5 went to a high team. So thats about 7%. So I suppose it's close.

But regardless. I guess we can still take the data from the top 100 signings of H over VH. With a small sample size, it's about 30%. But I find the fact that 3 of them were with just 1 high beating 2 VHs. That makes NO sense for it to happen that much.

Keep in mind that the considering tab reflects relative interest. When there are 3 battlers, it could be 35-35-30. We're going to see this more in the top 100 than anywhere else, because people aren't likely going to want to take those kinds of long odds farther down the line.

Yeah I was going to say it was 35-35-30. And that is silly in my opinion.

The added randomness can be heartbreaking but adds much needed reality. Having a player just go to the highest bidder is unrealistic. And not being able to predict an eighteen year old's behavior better mirrors college hoops recruiting. It is one of my favorite changes.

Posted by MyGeneration on 10/1/2016 9:32:00 AM (view original):
The added randomness can be heartbreaking but adds much needed reality. Having a player just go to the highest bidder is unrealistic. And not being able to predict an eighteen year old's behavior better mirrors college hoops recruiting. It is one of my favorite changes.

I don't mind the randomness of the signings. I get why they added it. I'm fine with it. And we're all going to have an opinion on how much randomness is too much.

The way it looks to me, if there is 1 VH vs 1 H then the H has a ~30% chance of signing. But if there are 2 VHs and 1 H then the H has a ~30% chance of signing. That part, to me, doesn't make any sense. The High's chance of winning should be drastically reduced if there are 2 VHs in play

So, here's another oddity that may goof up the data. Just had a kid sign with another school in Tark. I was moderate and they were very high. AFTER the recruit signed my interest level changed to high. So, do we know for sure the interest levels after signings are the same as before signings?

Posted by Benis on 10/1/2016 9:42:00 AM (view original):

Posted by MyGeneration on 10/1/2016 9:32:00 AM (view original):
The added randomness can be heartbreaking but adds much needed reality. Having a player just go to the highest bidder is unrealistic. And not being able to predict an eighteen year old's behavior better mirrors college hoops recruiting. It is one of my favorite changes.

I don't mind the randomness of the signings. I get why they added it. I'm fine with it. And we're all going to have an opinion on how much randomness is too much.

The way it looks to me, if there is 1 VH vs 1 H then the H has a ~30% chance of signing. But if there are 2 VHs and 1 H then the H has a ~30% chance of signing. That part, to me, doesn't make any sense. The High's chance of winning should be drastically reduced if there are 2 VHs in play

You're making up numbers based on a crazy small sample size and then getting made about it, stop! VH/H doesn't determine their chances of signing, their overall effort (which determines their chances of signing) determines VH/H. H could have anywhere from 1% to 30% would be my guess, but again, just a guess! H doesn't concretely mean 30%! It's a range based on effort!

Also, I've won some H over VH battles. Perhaps this is completely anecdotal, but it seems that when I win those, I have preferences that match up very well. Has anyone else seen this?

Posted by tkimble on 10/1/2016 9:58:00 AM (view original):

Posted by Benis on 10/1/2016 9:42:00 AM (view original):

Posted by MyGeneration on 10/1/2016 9:32:00 AM (view original):
The added randomness can be heartbreaking but adds much needed reality. Having a player just go to the highest bidder is unrealistic. And not being able to predict an eighteen year old's behavior better mirrors college hoops recruiting. It is one of my favorite changes.

I don't mind the randomness of the signings. I get why they added it. I'm fine with it. And we're all going to have an opinion on how much randomness is too much.

The way it looks to me, if there is 1 VH vs 1 H then the H has a ~30% chance of signing. But if there are 2 VHs and 1 H then the H has a ~30% chance of signing. That part, to me, doesn't make any sense. The High's chance of winning should be drastically reduced if there are 2 VHs in play

You're making up numbers based on a crazy small sample size and then getting made about it, stop! VH/H doesn't determine their chances of signing, their overall effort (which determines their chances of signing) determines VH/H. H could have anywhere from 1% to 30% would be my guess, but again, just a guess! H doesn't concretely mean 30%! It's a range based on effort!

I'm not mad* about this. I'm trying to understand what it is and if it makes sense. It started from what the dev chat response told us. I guess I misunderstood what they were saying. But regardless, I still want to gather data and see what it tells us.

I agree, it's a small sample size so far. But if others start doing the same then we can get an idea of what it looks like for battles and then we can decide if it makes sense or not. If it looks totally out of wack then we should let them know and see if they will change it. Otherwise we are completely guessing.

And yes, their overall effort determines their chance of singing but the VH and H indicate the zones of signing percentages. And yeah, we are guessing here but based upon everything we've seen so far and what Seble provided us in the Beta, a High interest doesn't have 1% to 30% chance.

Also, I was using 30% as an approximate. So around 30% chance. But again, if we all collect data from actual signings then we'll see how often a High beats a Very High. That will tell us approximately how likely it is that a H will beat a VH.

Posted by Benis on 10/1/2016 10:28:00 AM (view original):

Posted by tkimble on 10/1/2016 9:58:00 AM (view original):

Posted by Benis on 10/1/2016 9:42:00 AM (view original):

Posted by MyGeneration on 10/1/2016 9:32:00 AM (view original):
The added randomness can be heartbreaking but adds much needed reality. Having a player just go to the highest bidder is unrealistic. And not being able to predict an eighteen year old's behavior better mirrors college hoops recruiting. It is one of my favorite changes.

I don't mind the randomness of the signings. I get why they added it. I'm fine with it. And we're all going to have an opinion on how much randomness is too much.

The way it looks to me, if there is 1 VH vs 1 H then the H has a ~30% chance of signing. But if there are 2 VHs and 1 H then the H has a ~30% chance of signing. That part, to me, doesn't make any sense. The High's chance of winning should be drastically reduced if there are 2 VHs in play

You're making up numbers based on a crazy small sample size and then getting made about it, stop! VH/H doesn't determine their chances of signing, their overall effort (which determines their chances of signing) determines VH/H. H could have anywhere from 1% to 30% would be my guess, but again, just a guess! H doesn't concretely mean 30%! It's a range based on effort!

I'm not mad* about this. I'm trying to understand what it is and if it makes sense. It started from what the dev chat response told us. I guess I misunderstood what they were saying. But regardless, I still want to gather data and see what it tells us.

I agree, it's a small sample size so far. But if others start doing the same then we can get an idea of what it looks like for battles and then we can decide if it makes sense or not. If it looks totally out of wack then we should let them know and see if they will change it. Otherwise we are completely guessing.

And yes, their overall effort determines their chance of singing but the VH and H indicate the zones of signing percentages. And yeah, we are guessing here but based upon everything we've seen so far and what Seble provided us in the Beta, a High interest doesn't have 1% to 30% chance.

Also, I was using 30% as an approximate. So around 30% chance. But again, if we all collect data from actual signings then we'll see how often a High beats a Very High. That will tell us approximately how likely it is that a H will beat a VH.

We're starting to get an okay sample for VH vs H.

But the VH vs VH vs H seemed to be bothering you more, and our sample for that is super, super tiny.

If the Vh split chances and the lone h is not affected accordingly, it could swing the balance towards H.

You guys are right, the sample size is very small. If anyone has other examples they've seen from other worlds, post them up. But I thought 3 out of 4 is very high (no pun intended) for something that I personally think should be very rare. Maybe the next 20 instances will go towards the VH.