AbhishekLFC wrote:For me the solution is in the stage design, where the wide flat finishes are causing the problems.
This, this, this and this again. I have reported on a lot of races in pcm 18 and 20 and have always seen this as one of the biggest problems of lead-outs being an absolute mess, and this game version definitely doesn't seem to help our problem.
I am not saying every lead-out will always deliver the winner/top-3 when we can find some people to go back to 'normal' pcm finishes, but of course that also shouldn't be our goal.
I think the switch to long and extremely wide roads in (sprint) races were done the season I joined (or the one before), precisely in order to fix a sprinting 'problem' where top sprinters won too much, but by doing so, we created a new and seemingly way bigger problem. Pcm obviously wasn't made for 20+ 78 sprinters in one race to start with, but making the stages finish like this basically destroyed the AI. Like I said, only having 'normal' pcm roads won't magically make everyone happy, but the results seem to be way more normal/acceptable there. At least that's what I've noticed in the reports I have done myself.
If i recall the flat stage finishes were adjusted because in PCM 18 (?) only one leadout would ever survive turns due to poor AI handling solid rider models. Leading to strung out results where either you got on the right train or got nothing (something like that).
So if we're seeing now the opposite where leadouts are not strong enough, the stage design makes things worse.
But it's always been a case that sprints were comfortably the worst part of PCM. PCM10 you'd get glitched where the sprinter couldn't stop following the lead-out man. PCM12 was weird with ACC I think. PCM18 has the above, plus was that the edition with superpowered slipstream? PCM15 in my memory was a better sprint, but then it was other terrains it lacked in.
Heck, I spent years tied to Ewan suffering through it all and eventually gave up because the AI just can't make it work (tbf human controlled sprinting wasn't much better, far too exploitable because of the above mentioned issues). But again a big thing for me is the DB does nothing to help, only make problems worse.
Downside is testing an older PCM version requires converting the DB backwards, which I guess is quite a bit of time to do. I'd suggest coming back to this problem mid-season with a full overview of how all three divisions are going. SotD's above is relatively clear for PT, but in my head CT has worked reasonably well in the sprints so far, and this division has the most "normal" DB distribution. Anyway, just a few other thoughts.
Shocker, every version has a problem.
The solution that you always preach is the one logical and not nonsensical bullshit by pseudo intellectuals. If there is an 'issue' its the DB first and foremost and not the game, which Ive also always agreed with. A few years ago there was a large discussion about widescale stat reduction and even though 'its not fair', it sure is by far the best one if you actually 'care' and not pretend for moral high ground.
I would be perfectly fine just nuking stats to PCM levels
My impression is that the issue in PCM22 that is worse than PCM 20 is that lead outs have become even more toxic. So it isn't a question of the top sprinters doing badly it is a question of leadouts doing badly. I would have two suggestions to SotD if you continue your tracking:
-Extned it to the 4th and 5th ranked sprinter, my theory is they will tend to do better
-Consider flat in your ranking and weight AC over SP, and see how the best 5 riders do when you rank them by something like 50:20:30 FL/SP/AC
I did some testing on Qatar where I increased the FL of the top sprinters. It didn't change the results much but it did make them more consistent with SP/AC probably because FL and SP/AC were more correlated.
I am not sure if the big open roads is the whole problem, but certainly doesn't help. As the season goes on we will have more stages with traditional finishes. Guadeloupe for example has only 2 lane finishes I think. This isn't hard to fix if there is someone who is willing to spend some time editing stages.
The energy depletion is a big part what makes it random cause the game just randomly generates launch times and everyone burns out so quick. Been that way since PCM 18, but it's worse this year. Favorites are at the front and get trains, so they launch earlier and crack. I've noticed for a couple of years now that slight downhill finishes, where energy depletion is slower, tend to have more realistic results. The trains don't burn too fast, guys maintain their top speed for longer (which accents stat differences), and the cream rises to the top.
This of course doesn't account for when favorites just don't show up at all, but there's nothing we can do about that.
RIP Exxon Duke, David Veilleux, Double Feature, and Monster Energy
Could be interesting to test if Qatar with downhill sprints would change anything.
Obviously if we Can do small changes to get the results we are looking for more often it’s better than doing backtraces. I think we need to look at hills also. But I don’t have a top contender there so haven’t paid as much attention. Just feels weird in GC races too.
