Replies: 1 comment 9 replies
-
|
Also, maybe add where you can select which players may be questionable to play for an injury. That way the model can calculate that the player may not play at 100% returning from injury so the player may end up with maybe reduced minutes. |
Beta Was this translation helpful? Give feedback.
-
|
like i really think people dont realise how much complexity and time i put into this damn thing.... like there are soo many hidden features haha |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Nice, I do have some ideas to enhance based off what you mentioned. I'll provide in a new topic. |
Beta Was this translation helpful? Give feedback.
-
|
sounds good |
Beta Was this translation helpful? Give feedback.
-
|
i think i just need to write better docs or somthing??? like from your perspective what can i do make it easier to understand how to use the system??? like if added tour thing onboarding like that annoying shit i always skip and websites.... i could make somthing like that but that seems really annoying, i could make chat thing, and then give gpt the docs and all the info so gpt know how to use then people could chat and ask question and gpt could repsond hopefully informed of how the system works.... there is alot in the docs and there are like question mark icons that have alot of info on the live server too.... the local install is a bit more bare... because i have to be carefull about like the shop and there are difference with the live and local installs, its the exact same thing.... there same project but two different branches lets say.... although the live server is just live... its docent really have its own branch,.... oo this it i have my own local dev branch that is pretty much in sync with the live server..... idk how i got started on this. i was watching borat then i remember another thing.,.. idk man... honestly been working on this soo long still docent work soo open to any ideas.... also i think just messing with the system there are soo many options and ways to use it.... there is probably some value here and there... just have to find it..... i dont think vegas odds or predicting future are easy things....... how difficult idk yet to really get the system to point i want it to be at.... like ya know... would be great if it was just magic.... but we 4 years into this project.... i made the first folder of when i started writing stuff for this over 4 years ago. also i just want to be clear my background and expertise is in information systems. not basket ball, not vegas odds.... i dont bet on games, and am not a gambler.... not even really all that interested in betting. i just want to predict games.... and hopefully sell that system to other people to bet... but i dont want it to be scam... that why i work so hard to build system. Like it needs to be profitable to not be scam.... betting has limits. its not money printer. casinos just tell you no, sorry you can no long place bets, your a sharp better.... sooo you can make some money but not infinite..... soo i build system instead, not try to become gambler.... okay back to borat |
Beta Was this translation helpful? Give feedback.
-
i think i build last game 3 soon. last game 2 seems to have helped. then i need to start thinking about middle data, in between the short term momentum and long term season averages.... seems like things are doing a bit better, like there are sparks of it working. I am starting to only push new features to live server..... I plan on still support local installs. I think the overall idea that we have a base install that gets most things you already have now, and if you really want the cool new features then use live server.... that way i have a bit of a gap there.... and somtimes i may make big release with lots feature.
also notice change if you train with or without spread... like without spread you get some wild predictions, but they may be more accurate. And with spread model kinda follows what the spread does to some degree.... soo yea recommend trying models with and without spread... ohhh yea i want to build out drop all players.... ooo ill just make the slider go all the way to 0... then we can have model that dont dont use the top season average table, and then you could even not worry about removing injured players.... because it would just be team level stats and last game data.... not season averages with changing rosters......
okay i watch borat tonight.
Beta Was this translation helpful? Give feedback.
All reactions