0 calories

@MeisterHost of the game can make own rules & policies and can even ban anyone without noticing.. upto this part all is ok no matter who host. What is not good is "new" smurfs account with ability to immediatly host new game. Other players gladly will join without caring much who is the host as there is in general not much games hosted at peak time. I would like to offer devs team to implement some "smart logic" whoch would not allow host game from lobby for these "fresh accounts" and maybe additional conditions to make it harder you define... I don't have single proof mentioned person is Geriatrix (probably very first account was Cedric_O) but In general I never been banned by anyone else than him... if it helps..

I think most of the cases is when user having networking issue he will just rejoin with another provider mobile/wifi etc.. what has available to finish the game...

Just curious where did you get to the facts he is she?...

is it the auto-snipping script you "rumours" delivered to Stockfish? maybe share the replays what are you reffering to, let ffm2 or others to look into the activities. We don't want just spread rumors, lies etc.. right?

@ffm2 excellent work. I'm happy to see you confirm my words from the beggining. Yes hardware mouse gives advantage as any better device. Anyway definitely it is not auto-snipping with programmed smart logic of targeting. Now we must decide if this is cheating (as you pointed your own opinion that it is cheating) or this is standard way how to utilize better hardware we have on the market. What do you guys think? Should we ban specific mouse types? or gaming keyboards? (just imagine with gaming keyboard you also get 0.2ms advantage of poushing key) in general all better hw gives you little edge.

DO NOT RUN IT if you have larger replays directory! Your machine will get over busy and all you have running will crash. Script run all in paraler so it cant handle too many directorys = new process...! FORK BOMB - crash your PC. Exhaus memory triggier swap and kills everything

ohh there was multiple errors and the last i posted was misleading. I endup with this error: AttributeError: 'Legend' object has no attribute 'legend_handle and fixed it (here is why https://github.com/fastlmm/FaST-LMM/issues/48 and https://github.com/matplotlib/matplotlib/issues/28680) so if you face similar issue replace legend_handles with legendHandles thanks

Any idea what is wrong here? 2018-04-28_0001 Process Process-1: Traceback (most recent call last): File "/usr/lib/python3.10/multiprocessing/process.py", line 314, in _bootstrap self.run() File "/usr/lib/python3.10/multiprocessing/process.py", line 108, in run self._target(*self._args, **self._kwargs) File "/home/me/.local/share/0ad/replays/0.0.26/snipe_analyzer_platinum.py", line 31, in analyse_replay player_data = commands[0].split(',"PlayerData":[')[1] # crop unnecessary IndexError: list index out of range

This side gives me same results

@ffm Can you get overall statistic for lets say 100 games with me and see if these values are over-reached in all cases? I havent change it for year so I guess prove is in data - check and see .-) and i stand by my claims might be another aspec of these high values? is lag capable to recieve multiple timeframes within one or so?

@ffmmaybe consider adding on chart some output like CPS take all commands within every 1 minute period and count it. My test shows I reach average 20-22 CPS. Tested on this website https://clickspeeder.com/click-per-second/ but probably any other will work the same.

you simply want to avoid overkill (shots goes nowhere) - you have to move your mouse to spread attack to different units of course .-) so there will be many cases you just send attack to the ground (coordinates) :--) but if you zoom it correctly.. targeting is easier..

I have 11 mouse buttons, all of them have slightly different setup, sometimes I click button with single click, when I want snipping with extended multi-clicks with 25ms delay after. 200ms / 25ms = 8 action * number of clicks within this period (1-4 is my range) I agree you should see identical commands in reply for each action. Just get such mouse and simulate it yourself..

just run that python script ffm attached you can get it yourself

With community mod snipping is not that effective anymore, with vanilla it is significantly much more effective. 1000 click within minute probably wont be even registered by 0ad engine... you have 20msec one turn.. human is able to get from 5-15 click within this period depends how trained he is .. I would rather see Atrik recommendation to avoid "faster clicks" wins.. so engine would avoid it or get more automation into targeting..