smozoma

L26063 - from a reddit post

No problem, I DM'd you the spreadsheet and an idea for how to fix the problem (looks like they were using Time on Ice to solve the problem.. so all else being equal, the player with the most ice time would get the higher star!)

I must have removed the protection against not enough players in order to cram in more calculation logic Forgive me if I don't understand how this works anymore 11 years later:

Ohhh I wonder if that's a divide-by-zero on the save percentage. Doubt I'll fix it.. It was hard to cram the assembly instructions for my formula into the same number of bytes as the old formula, since mine is more complicated.

Check out line 950 in the .py file, it lists out the values used. If you want to add different logic rather than different values, that's waaaay more complicated. If you have any fixes to make in the file, such as the path, please add and share! The path stuff is around line 308, but assumes windows (backslashes)

I think it's 0.2. A -1 can turn a 30 into a 29. 30/5=6.0, 29/5 = 5.8.

So using a value of 4 gives the following odds for hot/cold? 2/8 odds of -1 (random values -4 and -3) 5/8 odds of 0 (random values from -2 to +2) 1/8 odds of +1 (random value of +3) And the +/-1 translates to +/-0.2 on the 0-6 scale (internally the game uses 0-30)

nice there is a 3-stars hack already which gives nice results for 5-minute period vs games: https://forum.nhl94.com/index.php?/topic/12181-tool-hack-applicator-aka-smozrom/

It's a good idea, but at a bad time. Live tournaments were taking off the past few years, but then the pandemic hit. I got super sick just from the 2019 King of 94 (being the scorekeeper everyone talked to, breathed on...), so there was no way I was hosting a tourney in 2020 or 2021. 2022 should be better now that most kids can get vaccinated (nearly half of Canada's known cases are in 0-20 year olds now. The other half is NHLers apparently ).

calling @seamor , he might know more.

I think you'd be the first here to try it. You have to enable Developer Mode on the xbox (requires paying $20) and then you can install it. Here are instructions: https://www.howtogeek.com/703144/how-to-install-the-retroarch-emulator-on-xbox-series-x-or-s/ I have no idea if netplay works between the xbox retroarch and PC/mac retroarch. In theory it should, but ¯\_(ツ)_/¯. We already have a problem playing Sega NHL'94 between PC and Mac. I suspect that while you're in Developer Mode your actual XBox games won't work and you have reboot into regular mode to play them.

Extra info about what your Elo rating means: It's a rating scheme where the rating difference between 2 players corresponds to a probability of winning. And points are transferred between the players according to that probability (much like betting odds), which makes the future ratings/odds more accurate. Over time if you play someone many times your rating difference will correspond very well with the percentage of times you won/lost according to the following relations: 0 point difference corresponds to (obviously) a 50% chance for either player to win. 100 point difference corresponds to a 64% chance of the higher-rated player winning. 200 -- 76% 300 -- 85% 400 -- 91% 500 -- 95% 50 -- 57% Our ratings were originally calibrated so that an average "B" league player has roughly a 1500 rating. (Thanks @Tickenest for his Elo calculations based on Classic league results!) An elite player with a 1900 rating will win ~91% of the time over a player with a 1500 rating (400 point difference). And a 1500 player will also win ~91% of the time over a 1100 rating player (400 point difference).

I was going to say, make sure there's not much colour in the background, but you're way ahead of me. Don't forget that greys count as a colours and there are 6 different greys possible, so maybe it would be better to use a hue shift to make the orange hat roughly the same colour as one of the team colours. Remove ads from boards. Like you said, only have a few core colours (white, black, maybe 3 colours in the uniforms..) I'd need to see some examples of images that don't convert well in order to realize why they are problematic. I'm hoping to update the software with the ability to give areas of focus so it doesn't optimize the ice or something. You could try brightening or darkening the image a little before running it through the program and it may actually make a really big difference in the result.

Note from facebook poster Scott Edward S: For Sega, you need to install RetroArch on the mini as well, because the default Genesis emulator is buggy and doesn't run NHL94 properly (lots of reports on here of backwards players, braindead goalies..).