Posted by johnsensing on 7/17/2017 3:18:00 PM (view original):

Posted by tecwrg on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

Seems like you're the one being deliberately obtuse, because you're arguing that 19% should = 0%, and 39% should = 0%.

Again, that's not how math works.

Read my last post again. I'm not arguing about math. If you don't understand that (and apparently you don't), you maybe should step away from the conversation.

Posted by johnsensing on 7/17/2017 3:23:00 PM (view original):

Posted by shoe3 on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

JS, remember the odds you see don't represent the difference in effort credit. The odds are stretched, so battles are always closer than they appear by looking at odds, and that's most exaggerated at the margins. A team has to be somewhere above 60% of the effort credit leader to be in signing range. A team that only just makes the cutoff isn't going to have the ~38% chance you'd expect if it was a straight correlation, though. It's going to be down around 20%. So a 50.1 to 49.9 battle in effort credit is stretched to 52-48. A 62-38 battle is stretched to something like 80-20.

The game is already leaning toward the favorite in recruiting battles.

What are you basing your first paragraph on? This is the first I've heard of any "odds-stretching." If a post-battle percentages reads 70% for Team A and 30% for Team B, I've been reading that as Team A had a 70% chance to win the recruit, and Team B had a 30% chance. If you're telling me that the actual percentages in my hypothetical are 80% for Team A and 20% for Team B (despite saying 70/30), that makes my point that this needs to be reworked even stronger, because that makes no sense.

Posted by tecwrg on 7/17/2017 4:20:00 PM (view original):

Posted by johnsensing on 7/17/2017 3:18:00 PM (view original):

Posted by tecwrg on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

Seems like you're the one being deliberately obtuse, because you're arguing that 19% should = 0%, and 39% should = 0%.

Again, that's not how math works.

Read my last post again. I'm not arguing about math. If you don't understand that (and apparently you don't), you maybe should step away from the conversation.

But it is about math. It's numbers: odds/percentages. That's math.

Sounds like your argument is that long shots should never win. That doesn't make sense. There's lots of reasons why long shots should occasionally win.

Let's say it's a two way battle between Indiana (81%) and Iowa (19%). Indiana should win most of those battles, according to the math that you seem to dislike. But occasionally, some recruit might decide they don't want to play for Bobby Knight because he sometimes throws chairs and chokes his players. Maybe their girlfriend is going to Iowa. Maybe they like sitting in cornfields while pondering the merits of a 2-3 zone.

Posted by tecwrg on 7/17/2017 4:20:00 PM (view original):

Posted by johnsensing on 7/17/2017 3:18:00 PM (view original):

Posted by tecwrg on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

Seems like you're the one being deliberately obtuse, because you're arguing that 19% should = 0%, and 39% should = 0%.

Again, that's not how math works.

Read my last post again. I'm not arguing about math. If you don't understand that (and apparently you don't), you maybe should step away from the conversation.

But it is about math. It's numbers: odds/percentages. That's math.

Sounds like your argument is that long shots should never win. That doesn't make sense. There's lots of reasons why long shots should occasionally win.

Let's say it's a two way battle between Indiana (81%) and Iowa (19%). Indiana should win most of those battles, according to the math that you seem to dislike. But occasionally, some recruit might decide they don't want to play for Bobby Knight because he sometimes throws chairs and chokes his players. Maybe their girlfriend is going to Iowa. Maybe they like sitting in cornfields while pondering the merits of a 2-3 zone.

Posted by tecwrg on 7/17/2017 4:20:00 PM (view original):

Posted by johnsensing on 7/17/2017 3:18:00 PM (view original):

Posted by tecwrg on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

Seems like you're the one being deliberately obtuse, because you're arguing that 19% should = 0%, and 39% should = 0%.

Again, that's not how math works.

Read my last post again. I'm not arguing about math. If you don't understand that (and apparently you don't), you maybe should step away from the conversation.

But it is about math. It's numbers: odds/percentages. That's math.

Sounds like your argument is that long shots should never win. That doesn't make sense. There's lots of reasons why long shots should occasionally win.

Let's say it's a two way battle between Indiana (81%) and Iowa (19%). Indiana should win most of those battles, according to the math that you seem to dislike. But occasionally, some recruit might decide they don't want to play for Bobby Knight because he sometimes throws chairs and chokes his players. Maybe their girlfriend is going to Iowa. Maybe they like sitting in cornfields while pondering the merits of a 2-3 zone.

Posted by pkoopman on 7/17/2017 4:24:00 PM (view original):

Posted by johnsensing on 7/17/2017 3:23:00 PM (view original):

Posted by shoe3 on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

JS, remember the odds you see don't represent the difference in effort credit. The odds are stretched, so battles are always closer than they appear by looking at odds, and that's most exaggerated at the margins. A team has to be somewhere above 60% of the effort credit leader to be in signing range. A team that only just makes the cutoff isn't going to have the ~38% chance you'd expect if it was a straight correlation, though. It's going to be down around 20%. So a 50.1 to 49.9 battle in effort credit is stretched to 52-48. A 62-38 battle is stretched to something like 80-20.

The game is already leaning toward the favorite in recruiting battles.

What are you basing your first paragraph on? This is the first I've heard of any "odds-stretching." If a post-battle percentages reads 70% for Team A and 30% for Team B, I've been reading that as Team A had a 70% chance to win the recruit, and Team B had a 30% chance. If you're telling me that the actual percentages in my hypothetical are 80% for Team A and 20% for Team B (despite saying 70/30), that makes my point that this needs to be reworked even stronger, because that makes no sense.

This was discussed at length with seble in beta. The final odds are not representing where effort credit is. They are stretched, in favor of the team in the lead. As in my (shoe3's) example, if you are beating someone 50.1 to 49.9 in effort credit, when the battle is over, you'll see the odds at 52-48, not 50-50, or 50.1 to 49.9. 54-46 in effort credit gets stretched to something like 59-41 when the recruit makes a decision. In essence, WIS is putting a finger on the scale in favor of the team in the lead. It's slight when the battle is close - not much difference between a true toss up at 50-50 and a 52-48 probability roll. But it's significant when you get to the battles that are true upsets, where a battle that is 62-38 (the minimum effort credit for signing range is roughly 60% of the leader, which was as much info we could squeeze out of seble) in terms of real effort credit turns out as 80-20 when you roll.

So if you are in a two way battle for a recruit, say the leader has 620 points of effort credit (this is hypothetical, because we don't know the relative scale of the types of effort) and the trailer has 380 points. This is roughly where the cutoff is for the trailer to be considered. The roll is not going to be 62-38. It's going to be something like 80-20 (I haven't seen a bigger gap in odds, so I'm working under the assumption that 80-20 is the extreme).

Posted by johnsensing on 7/17/2017 5:24:00 PM (view original):

Posted by pkoopman on 7/17/2017 4:24:00 PM (view original):

Posted by johnsensing on 7/17/2017 3:23:00 PM (view original):

Posted by shoe3 on 7/17/2017 11:33:00 AM (view original):

Posted by johnsensing on 7/17/2017 11:19:00 AM (view original):

Posted by tecwrg on 7/17/2017 9:27:00 AM (view original):
You all realize that teams with a 19% chance of winning a recruiting battle should win roughly 19% of those battles. You guys sound like they should win 0% of those recruiting battles.

That's not how math works.

Unless you have some empirical evidence that underdogs are winning battles at a statistically significant higher rate than they should, then there's nothing really to see here. My guess is that such evidence does not exist, at least in a way measurable to the HD user community, because neither the "favorites" who are winning the lopsided battles, nor the underdogs who are losing the lopsided battles, are making noise in the forums the way the favorite/losers are.

Are you being deliberately obtuse? I don't think anyone is arguing the math -- I'm arguing policy/gameplay. Of course as things are currently set up, someone who has a 19% chance will win 19% of the time -- my argument is that it's a poor way to set up the game, and the game should be changed so that a 19% in a three-man battle has 0% chance to win (I'm not sure where the cutoff should be in a three-man battle -- 25%? 28?). What I am advocating is that for two-person battles, if one side hasn't put in enough effort to get to a 40% probability, they should have 0% chance. That's a better gameplay mode, IMO, because then you only have losses when it's a true tossup, or at least pretty close (which is really the way it works in real life, too, which is an added benefit).

JS, remember the odds you see don't represent the difference in effort credit. The odds are stretched, so battles are always closer than they appear by looking at odds, and that's most exaggerated at the margins. A team has to be somewhere above 60% of the effort credit leader to be in signing range. A team that only just makes the cutoff isn't going to have the ~38% chance you'd expect if it was a straight correlation, though. It's going to be down around 20%. So a 50.1 to 49.9 battle in effort credit is stretched to 52-48. A 62-38 battle is stretched to something like 80-20.

The game is already leaning toward the favorite in recruiting battles.

What are you basing your first paragraph on? This is the first I've heard of any "odds-stretching." If a post-battle percentages reads 70% for Team A and 30% for Team B, I've been reading that as Team A had a 70% chance to win the recruit, and Team B had a 30% chance. If you're telling me that the actual percentages in my hypothetical are 80% for Team A and 20% for Team B (despite saying 70/30), that makes my point that this needs to be reworked even stronger, because that makes no sense.

This was discussed at length with seble in beta. The final odds are not representing where effort credit is. They are stretched, in favor of the team in the lead. As in my (shoe3's) example, if you are beating someone 50.1 to 49.9 in effort credit, when the battle is over, you'll see the odds at 52-48, not 50-50, or 50.1 to 49.9. 54-46 in effort credit gets stretched to something like 59-41 when the recruit makes a decision. In essence, WIS is putting a finger on the scale in favor of the team in the lead. It's slight when the battle is close - not much difference between a true toss up at 50-50 and a 52-48 probability roll. But it's significant when you get to the battles that are true upsets, where a battle that is 62-38 (the minimum effort credit for signing range is roughly 60% of the leader, which was as much info we could squeeze out of seble) in terms of real effort credit turns out as 80-20 when you roll.

So if you are in a two way battle for a recruit, say the leader has 620 points of effort credit (this is hypothetical, because we don't know the relative scale of the types of effort) and the trailer has 380 points. This is roughly where the cutoff is for the trailer to be considered. The roll is not going to be 62-38. It's going to be something like 80-20 (I haven't seen a bigger gap in odds, so I'm working under the assumption that 80-20 is the extreme).

I had never seen this before -- I left the beta in disgust early. Has this been checked with actual battles? Did seble say why he implemented the "odds stretching"? It seems really foolish to have a system where I put in 55% of effort credit, and the other team put in 45%, but neither one of us has the actual odds of winning that: (a) our respective effort levels would seem to dictate that we would; or (b) that the "considering" screen actually says. It also makes no sense whatsoever to implement a system that explicitly is designed to give worse teams (or lower-prestige teams, if you like that better) a better chance, and then give the leading team on effort, which will more often than not be a higher-prestige team, a thumb on the scale.

More evidence for my thesis that the entire 3.0 system is a botch, I suppose.

Posted by Benis on 7/17/2017 5:30:00 PM (view original):
I didn't use the term 'stretching the odds' since it's a little confusing but that is what Seble said. I used a bonus applied which is another way Seble used to describe it and probably a little easier to understand.

Posted by johnsensing on 7/17/2017 5:38:00 PM (view original):

Posted by Benis on 7/17/2017 5:30:00 PM (view original):
I didn't use the term 'stretching the odds' since it's a little confusing but that is what Seble said. I used a bonus applied which is another way Seble used to describe it and probably a little easier to understand.

But why did he do it that way? That makes no sense to me at all.

Posted by johnsensing on 7/17/2017 5:38:00 PM (view original):

Posted by Benis on 7/17/2017 5:30:00 PM (view original):
I didn't use the term 'stretching the odds' since it's a little confusing but that is what Seble said. I used a bonus applied which is another way Seble used to describe it and probably a little easier to understand.

But why did he do it that way? That makes no sense to me at all.

Posted by zorzii on 7/17/2017 3:16:00 PM (view original):

Posted by MikeT23 on 7/17/2017 3:01:00 PM (view original):

Posted by darnoc29099 on 7/17/2017 1:48:00 PM (view original):
And before I get slaughtered on the "realism" aspect of the game the scenario outlined above is another tough sell to a new user. The game isn't going to be 100% realistic but there are lots of issues that just make much sense.

I recognize losing a 70/30 battle is frustrating(I've lost one of similar percentages) but you also win one on occasion(I've won one at 44, IIRC).

This is NOT something that keeps new users away. Hell, they probably have no idea because I see, in these forums all the time, "Where are you guys getting these percentages?"

They don't battle. D3 is a wait and reallocate ap long void.