Difference between revisions of "TMVL NMR Design"
(Created page with "This section will be used for developing a design when we need an automated NMR program to create lineup orders and recruiting orders. === Lineup Orders === ''Add ideas below...") |
|||
(4 intermediate revisions by 3 users not shown) | |||
Line 2: | Line 2: | ||
=== Lineup Orders === | === Lineup Orders === | ||
+ | We have an existing NMR program now. Essentially this will take in the number of players at a position and randomly assign one to the starting lineup. This runs once for each match. It does not try to take a full session view and say "hey, if its a 3-match session, I really should only allow a player to play 2 matches". So it can result into both some FIT +1 and some FIT -1 scenarios that could get uglier as the season progresses on the negative FIT side. | ||
+ | |||
+ | So we really need a more thoughtful program. Such as preventing a player from EVER playing more than 2 matches a session. How good or bad we make the NMR lineups is open to discussion but we could use something like "overall" as a guide to determine how much a player plays in a 2, 3, or 4 match session. | ||
+ | |||
+ | I'd prefer not to just make NMR lineups terrible (i.e. play the WORST of 2 setters 2/3 matches and the BEST in 1/3 matches a session). But maybe some randomness to that. | ||
+ | |||
+ | Also perhaps an emphasis on more playing time/XPs for Freshman. So really I don't have a clear direction, just some background. | ||
+ | |||
+ | |||
''Add ideas below'' | ''Add ideas below'' | ||
+ | |||
+ | I like your ideas Al. It feels like NMR teams should be ‘development’ sides so that a new coach can walk into a reasonable setup. For me that would mean prioritising XP over results when choosing line-ups. I’d agree with only allowing players to appear in 2 matches per session to keep FIT neutral but perhaps focus more T7 players on winnable games (home before neutral before away games) to maximise CPs and/or rank competitions in order of importance. | ||
+ | |||
+ | I don't know how easy it would be to code but some players only need to play one game per session to get the XP they need e.g 3* L and MB but a 5* Jr OH like Brehme at Columbia needs two games per session and some. So maybe some kind of system to prioritise who always gets two games. With D or worse potential they don't need to play (4th MB and OH in a 3 game session). | ||
+ | |||
+ | As for the lineup, say in a 3 game session decide which L gets 2 games (if both need just one then the stronger gets the second) and randomly assign to each game. | ||
* | * | ||
* | * | ||
+ | === CP Orders === | ||
+ | This would be a focus on Skills and not any usage of CP for FIT | ||
+ | |||
+ | ''Add ideas below'' | ||
+ | |||
+ | Above idea makes sense to me. Perhaps, as a development side, also prioritise coaching based first on age - with youngest first - and then with impact according to the equations I recall seeing somewhere. I think that would mean Setters SET first, then MB BLOCK, you’ll know the rest! | ||
+ | |||
+ | I would consider prioritising Potential grades, so a potential A Sophomore rather than a potential C freshman. | ||
=== Recruiting Orders === | === Recruiting Orders === | ||
''Add ideas below'' | ''Add ideas below'' | ||
+ | |||
+ | Nothing clever to add here. Suggest aiming for four players per season with one a free 3*. Which positions to target would depend on requirements (maybe just a like for like for who’s retiring?) For the competitive recruiting, as there’s three, how about just aiming for 1 each at 5*, 4*, 3*? Bids to be a proportion of budget, with a random element, aiming to have 1000 RP in the bank for the start of the next season? | ||
* | * | ||
* | * |
Latest revision as of 15:26, 13 September 2024
This section will be used for developing a design when we need an automated NMR program to create lineup orders and recruiting orders.
Lineup Orders
We have an existing NMR program now. Essentially this will take in the number of players at a position and randomly assign one to the starting lineup. This runs once for each match. It does not try to take a full session view and say "hey, if its a 3-match session, I really should only allow a player to play 2 matches". So it can result into both some FIT +1 and some FIT -1 scenarios that could get uglier as the season progresses on the negative FIT side.
So we really need a more thoughtful program. Such as preventing a player from EVER playing more than 2 matches a session. How good or bad we make the NMR lineups is open to discussion but we could use something like "overall" as a guide to determine how much a player plays in a 2, 3, or 4 match session.
I'd prefer not to just make NMR lineups terrible (i.e. play the WORST of 2 setters 2/3 matches and the BEST in 1/3 matches a session). But maybe some randomness to that.
Also perhaps an emphasis on more playing time/XPs for Freshman. So really I don't have a clear direction, just some background.
Add ideas below
I like your ideas Al. It feels like NMR teams should be ‘development’ sides so that a new coach can walk into a reasonable setup. For me that would mean prioritising XP over results when choosing line-ups. I’d agree with only allowing players to appear in 2 matches per session to keep FIT neutral but perhaps focus more T7 players on winnable games (home before neutral before away games) to maximise CPs and/or rank competitions in order of importance.
I don't know how easy it would be to code but some players only need to play one game per session to get the XP they need e.g 3* L and MB but a 5* Jr OH like Brehme at Columbia needs two games per session and some. So maybe some kind of system to prioritise who always gets two games. With D or worse potential they don't need to play (4th MB and OH in a 3 game session).
As for the lineup, say in a 3 game session decide which L gets 2 games (if both need just one then the stronger gets the second) and randomly assign to each game.
CP Orders
This would be a focus on Skills and not any usage of CP for FIT
Add ideas below
Above idea makes sense to me. Perhaps, as a development side, also prioritise coaching based first on age - with youngest first - and then with impact according to the equations I recall seeing somewhere. I think that would mean Setters SET first, then MB BLOCK, you’ll know the rest!
I would consider prioritising Potential grades, so a potential A Sophomore rather than a potential C freshman.
Recruiting Orders
Add ideas below
Nothing clever to add here. Suggest aiming for four players per season with one a free 3*. Which positions to target would depend on requirements (maybe just a like for like for who’s retiring?) For the competitive recruiting, as there’s three, how about just aiming for 1 each at 5*, 4*, 3*? Bids to be a proportion of budget, with a random element, aiming to have 1000 RP in the bank for the start of the next season?