alright, thanks guys for all the info on whether your stuff was working! sounds like still having problems. i am going to confirm the dynasty list data collector still has problems too, when i get home later to my computer that has access (technically, i have access from here too, i just don't know what to connect to and all that crap is saved on my home pc). if it does, which i am expecting, then i'll submit a ticket and see if we can get them looking into it.

Posted by gillispie on 3/22/2022 4:02:00 PM (view original):

Posted by mlitney on 3/22/2022 3:12:00 PM (view original):
I switched up the zone formulas when I first started using the spreadsheet years ago. It really bugged me that the entire team was averaged together instead of using the actual averages (PG-SG-SF, PF-C in 3-2 and PG-SG, SF-PF, C in 2-3).

Overall I've gotten so much use out of this spreadsheet. It's now over 10 tabs for each team after all of the things that have been added over time.

Thanks to hughes and cubcub for all your work putting this together!

the entire team is averaged together, the 'actual averages' you are talking about are the sets of players who share the same defensive ability equation

Right .. but in my spreadsheet, I also look at each position and compare each to the same position on the other team.

If playing zone .. for Defense, instead of comparing my PF1's offensive rating to their PF1 defensive rating .. I should compare my PF1 to an average of their PF+C (or PF+SF) , etc.

Posted by hughesjr on 3/22/2022 3:55:00 PM (view original):
Yep .. I had issues again this morning // my issues were in Allen and it happened just after rollover.

Maybe it has something to do with web page caching from a proxy server. I did not try to log any forwards that happen, but it is common practice for websites to cache active pages for period of time to reduce load. At roll over they would need to regenerate all those cached pages.

Also, I am not sure how well goggle sheets (or excel for that matter) will do auto forwarding .. and things are auto-forwarded from http to https. Maybe the process is different on rollover day.

Everything SEEMS to be working now for me again.

yeah... i think with the inner workings of excel or google sheets or whatever, it would be a tough angle for you guys, trying to get WIS to sort out why the sheet won't connect. what i do is more straight forward, less layers of opaque crap between me and them. so i am kinda thinking it should be easier for me to try to get something out of them, than for you guys?

Posted by hughesjr on 3/22/2022 4:09:00 PM (view original):

Posted by gillispie on 3/22/2022 4:02:00 PM (view original):

Posted by mlitney on 3/22/2022 3:12:00 PM (view original):
I switched up the zone formulas when I first started using the spreadsheet years ago. It really bugged me that the entire team was averaged together instead of using the actual averages (PG-SG-SF, PF-C in 3-2 and PG-SG, SF-PF, C in 2-3).

Overall I've gotten so much use out of this spreadsheet. It's now over 10 tabs for each team after all of the things that have been added over time.

Thanks to hughes and cubcub for all your work putting this together!

the entire team is averaged together, the 'actual averages' you are talking about are the sets of players who share the same defensive ability equation

Right .. but in my spreadsheet, I also look at each position and compare each to the same position on the other team.

If playing zone .. for Defense, instead of comparing my PF1's offensive rating to their PF1 defensive rating .. I should compare my PF1 to an average of their PF+C (or PF+SF) , etc.

well, i was talking about the sim engine. which maybe is clear to you but i wasn't 100% sure we are talking about the same thing here. i think mlitney has the common misconception, that in HD, zone players are averaged like pg/sg in 2-3 or sf/pf/c in 2-3. but really, all 5 players are averaged together for every shot, and the pg/sg in the 2-3 just share the same defensive formula, which means you can flip your pg/sg in a 2-3 without directly impacting the defense. where as flipping a sg/sf in a 2-3, would impact the defense, even though folks are averaged together.

anyway, i am not familiar with the details of your sheet at all, but when you say... 'for Defense, instead of comparing my PF1's offensive rating to their PF1 defensive rating .. I should compare my PF1 to an average of their PF+C (or PF+SF) , etc.'

for the purpose of comparing your pf1's offensive rating to their defensive rating, you 'should' compare to either the whole team, or nothing, because comparing your PF1 offense to their PF1 defense is inherent nonsensical.

i think anyway? but i thought mlitney was saying... what you ARE doing, is comparing to the whole team. which IMO, makes sense. what he says he changed it to, sounds to me like you what you just said is how it works. which IMO, does not make sense (as a thing to do).

Gill, I was talking about the portion of the spreadsheet that calculates a single player's offensive rating for a single game. This formula then ties into your team's overall odds of winning.

For example, to find the opposing PFs offensive chances against my 3-2 zone, I would look at the average defensive ratings of my PF-C. I think the default formula just took the average defensive ratings of the entire team to calculate how a single player would be defended when playing a zone instead of breaking it down into the groupings.

Hopefully that makes sense and please let me know if I'm doing it incorrectly lol.

Posted by mlitney on 3/22/2022 9:47:00 PM (view original):
Gill, I was talking about the portion of the spreadsheet that calculates a single player's offensive rating for a single game. This formula then ties into your team's overall odds of winning.

For example, to find the opposing PFs offensive chances against my 3-2 zone, I would look at the average defensive ratings of my PF-C. I think the default formula just took the average defensive ratings of the entire team to calculate how a single player would be defended when playing a zone instead of breaking it down into the groupings.

Hopefully that makes sense and please let me know if I'm doing it incorrectly lol.

i think it makes sense, and i think that is doing it incorrectly.

to find the opposing PF's offensive chances against your 3-2 zone, you need to look at the defensive ratings of all 5 players, who contribute to the defense of every shot. comparing to the average defensive ratings of your PF-C is in a way very similar to looking at the defensive rating of the PF alone - its not like what you are looking at is totally irrelevant, but it doesn't reflect the reality of zone defense.

I think I see what you're saying gill. So if my team's ratings looked like this:

Pos/Ath/Spd/Def
PG/60/90/80
SG/80/80/75
SF/85/65/90
PF/90/30/90
C/75/15/80

In a 3-2 zone, they would be grouped as PG/SG/SF and PF/C to find each player's individual defensive ratings. PG/SG/SF would be 75 ath, 78 spd, and 81 def for all 3 players. PF/C would be 83 ath, 23 spd, and 85 def for each player's individual defensive ratings.

But when looking at an opposing player's chances of scoring, we'd look at the entire team average, which would be:

Pos/Ath/Spd/Def
PG/75/78/81
SG/75/78/81
SF/75/78/81
PF/83/23/85
C/83/23/85

And those are the numbers that the sim would use to determine the chance of a successful shot. Is that basically what you're saying? This is a bit confusing because I thought that opposing players were "guarded" by a particular group. So the opposing PF would be guarded by my PF/C average ratings. Would the PG/SG/SF ratings make a difference as well when considering a low post shot by a PF?

Also, the sheet isn't working for me again this morning. It didn't seem to work yesterday morning as well. Maybe its a new morning thing?

Posted by mlitney on 3/23/2022 10:22:00 AM (view original):
I think I see what you're saying gill. So if my team's ratings looked like this:

Pos/Ath/Spd/Def
PG/60/90/80
SG/80/80/75
SF/85/65/90
PF/90/30/90
C/75/15/80

In a 3-2 zone, they would be grouped as PG/SG/SF and PF/C to find each player's individual defensive ratings. PG/SG/SF would be 75 ath, 78 spd, and 81 def for all 3 players. PF/C would be 83 ath, 23 spd, and 85 def for each player's individual defensive ratings.

But when looking at an opposing player's chances of scoring, we'd look at the entire team average, which would be:

Pos/Ath/Spd/Def
PG/75/78/81
SG/75/78/81
SF/75/78/81
PF/83/23/85
C/83/23/85

And those are the numbers that the sim would use to determine the chance of a successful shot. Is that basically what you're saying? This is a bit confusing because I thought that opposing players were "guarded" by a particular group. So the opposing PF would be guarded by my PF/C average ratings. Would the PG/SG/SF ratings make a difference as well when considering a low post shot by a PF?

sort of. your big intermediate step, to average the pg-sg-sf together, is superfluous. it is a step that is not technically incorrect in any way, but it accomplishes nothing.

here is how i envision it:
Pos/Ath/Spd/Def
PG/60/90/80
SG/80/80/75
SF/85/65/90
PF/90/30/90
C/75/15/80

then there is a formula to convert each player to a score. the defenses of the 5 players, they are never going to be directly averaged. the defensive scores however, will be. so let's say
fn defpg(ath, spd, def) = 1*def + .7*ath + .5*spd
fn defc(ath, spd, def) = 1*def + .8*ath + .2*spd

then the def score for each of the 5 players would be as follows:
defpg (PG/60/90/80) => 167 (this is, hopefully, 80 + 60 * .7 + 90 * .5)
defpg (SG/80/80/75) => 171
defpg (SF/85/65/90) => 187
defc (PF/90/30/90) => 168
defc (C/75/15/80) => 143

then your average defense would be the average of those 5, basically, 836/5 = 167.2

alright - so that is the thrust of it. with that sort of approach, hopefully you can see how the numbers work out the same regardless of your intermediate averaging step for the pg-sg-sf and pf-c. they use the same equations, so what you are doing there hurts nothing, but gains nothing. i want to make sure we are square on the core of the misconception before moving on. i am going to move on but you might want to ignore this next part until we are square on the rest of it.

so, the other part of this (i don't have more parts in my pocket), is that the defensive functions relate to the distance from the basket. everyone knows the above function is a simplification, its not fn defpg (ath, spd def), we all know its more like fn defpg (ath, spd, def, blk, def iq, fatigue). but what you may not realize is, its really more like this (fn is just a function short hand... transforms inputs into outputs, more or less).

fn defpg (ath, spd, def, blk, def iq, fatigue, distance from basket).

that last one is a bit of a doozy, but it makes a lot of sense if you sit on it for a minute. your pgs ability to defend 3 feet from the basket and 25 feet is going to be very different, right? so that is what that is about.

so, now that you have converted your way of thinking from, pg-sg-sf in a 3-2 zone, to the whole team... let me just completely muddy the waters. at this point we are leaving sort of known science and entering the realm of random gillispie contemplations. here is what i assume is really happening. when you have a guy defending a 25 footer, his speed is going to be a much greater factor in the defensive equation, than it is at 5 feet. now, at 25 feet, all 5 players are going to contribute - but who is really bringing the speed? your 1-2-3 right... and even if your 4-5 brought the speed, likely their defensive equations do not weight speed THAT heavily at any distance.

similarly, at 3 feet, blk is probably pretty critical. and everyone's block could contribute, but at 3 feet, the pf-c formula probably has much higher weight on blk, than the pg-sg-sf formula does. so even if you had 100 blk everywhere, its probably the 4-5 doing most of the work.

so i kinda think its probably fair to say, that against a 99 per, 1 lp, +2 scorer, the pg-sg-sf have an outsized influence on his scoring. i suspect its still pretty far from the pf-c doing nothing there, though, because i imagine def is highly relevant to all shots, and all players can bring 95 def you know? and also, for a 99 lp, 1 per center on -2, the pf/c are probably particularly relevant. but again, i doubt they are anywhere close to 100% of the contribution. maybe instead of 40%, they are 50-60%, who knows? i am not a zone expert - i consider myself to know next to nothing about zone really. so i can't really weigh in on how this all shakes out. some day i will go back and get my 2nd zone title so i can say i have multiple titles in every set, maybe, its on my bucket list to run a real d1 zone program one day, and to actually make sense of the damn thing. but if i do its at least 1-2 years out from even starting. so probably never. but a man can dream?

Wow, enlightening. I didn't know the code was so complex. I just figured it was a bunch of nested if statements written on a graphing calculator in 2000. Just kidding. But if your theory on shot distance is accurate, that would be more subtlety than I realized.

I just don't think it's realistic to capture those types of subtleties within a game planning spreadsheet as there's no way to easily discern the exact distance of each player's shots for the season. If you could somehow export all of that info to a spreadsheet, then we'd be in business.

Posted by mlitney on 3/23/2022 12:39:00 PM (view original):
Wow, enlightening. I didn't know the code was so complex. I just figured it was a bunch of nested if statements written on a graphing calculator in 2000. Just kidding. But if your theory on shot distance is accurate, that would be more subtlety than I realized.

I just don't think it's realistic to capture those types of subtleties within a game planning spreadsheet as there's no way to easily discern the exact distance of each player's shots for the season. If you could somehow export all of that info to a spreadsheet, then we'd be in business.

sincerely, i hope that never happens. i don't think its good for a game planning sheet to be THAT good. our tools beating the eye test of a high end coach, its an existential threat to HD and all similar games, in my opinion.

also, just to kinda be clear - the part about the averaging of the 5 players evenly, like in a mean style average, is also gillispie theory. it comes from seble that all 5 players are averaged together on every shot, in an extensive and detailed enough conversation that i'm fairly confident there's not a communication breakdown there causing us to receive a greatly different message than what was being sent by CS, if you will...

the part about the distance from basket, being part of the formula, is from seble.

the part of speed being worth more at greater distance, block worth less, that is me pontificating, as is the conclusion that there is 'essentially' a weighting by position as a result, even if it is highly nebulous and nearly impossible to quantify. but it does seem like common sense, no?

Posted by gillispie on 3/23/2022 12:58:00 PM (view original):

Posted by mlitney on 3/23/2022 12:39:00 PM (view original):
Wow, enlightening. I didn't know the code was so complex. I just figured it was a bunch of nested if statements written on a graphing calculator in 2000. Just kidding. But if your theory on shot distance is accurate, that would be more subtlety than I realized.

I just don't think it's realistic to capture those types of subtleties within a game planning spreadsheet as there's no way to easily discern the exact distance of each player's shots for the season. If you could somehow export all of that info to a spreadsheet, then we'd be in business.

sincerely, i hope that never happens. i don't think its good for a game planning sheet to be THAT good. our tools beating the eye test of a high end coach, its an existential threat to HD and all similar games, in my opinion.

Yes, my eye test is much better than my spreadsheet. It's a reason why HD is such a good game. Relying too heavily on my spreadsheet will definitely prevent development as a coach down the road.

Yeah, I agree. I wouldn't want to have the game figured out in that manner. To be honest, if I'm game planning for an important match, I'll usually do more research and often don't even use the recommended defensive positioning. The part that I really enjoy is the roster section. Just having a view of the opposing team ordered by position and starter/backup with all their ratings really helps me get an idea of what they're doing. Their strengths and weaknesses. How to make adjustments to get the most out of my game plan. It's a great tool for getting better at this game.

Oh wow sheet just loaded for literally the first time for my Naismith team, have been checking on and off. Naismith was by far my biggest problem, I'm in every 2 a day world and even if there was a problem there I usually could get it to work shortly after. Just my personal experience, also Cub again thanks for the sheet it is absolutely an incredible way to learn the game and of course Hughes as well for his sheet.

Also Cub in terms of resources that help the game, if you ever feel like breaking out a microphone and screen recording again, if you provide that stuff I'll edit the youtube videos and get them uploaded nicely with thumbnails. Those videos were so key to me when I first started and I still go back to check to see if you dropped any info that I now understand because this game is so complex and deep. I would make some myself but I just don't have that knowledge yet.