Hi all,

To summarize and in order to guide newcomers, the current medal holders [@chessjerk @MasterSystemMarceau @JFred99] + @Nimn_One and @AlphaNerd have all agreed to run the game at 60Hz when it comes to medal runs. It's not a rule from the moderation team, it's just something we decided to do between us.

Remember, if you're driving at 293km/h at 60Hz (the full speed in OutRun SMS), it means you're driving at 83% of the speed at 50Hz and therefore driving at 243km/h, which makes it a lot easier and not competing on the same level as other runners.

Again, this is not a rule from the moderation team. Feel free to upload runs at 50Hz but we would like to kindly unofficially ask you to upload runs at 60Hz if you're going for a medal.

Thank you everyone and long live this awesome community!

Hi, interested outsider from Super Hang-On. I volunteered to take a look into the SMS Outrun timing as Chessjerk and MSM have told me it feels wonky. I'll update when I have more info, but the first important thing is that the game seems to be unable to keep to its intended speed of processing, viz, a sort of lag affects things.

Example: stay still at the start line and watch the timer tick. It ticks after 61 then always 60 frames, regardless of PAL/NTSC. HOWEVER. If you accelerate away normally, that behaviour immediately breaks: timer seconds take 78 65 67 74 75 82 etc frames for NTSC, but seem to take 75 evenly for PAL, at least on the opening segment. This would mean that yes, the game is running slower at PAL, but that the game has more time per cycle to handle things. That would corroborate with my previous experience in Lemmings, where real time is used because the IGT has inconsistency.

I think a good next step would be for me to make a lua script that will print the frames between timer ticks, and get a skilled runner to do a run with Bizhawk's frame counter on, at PAL, and then at NTSC, and verify these initial findings. It's likely but not certain IMO that game timer behaviour will be similarly hiccup-prone, but you may prefer to ask someone experienced at making TASes to delve deeper as I may hit a skill ceiling. I'll post my script here when ready and get feedback from there, I can try some test runs myself but a top runner using PAL can give better info on what they feel as for lag of any sort.

For now I recommend that top runs be kept to NTSC to maintain your current ruleset and consistency, and to keep using RTA as a secondary info point, I would also encourage you to discuss your wishes for that in future because it may mean lag reduction strats could be found/change the order of current board runs.

I have been so far unable to get an accurate script working but I did an experiment with fixing track pos, gear, and speed through runs. When I did that, PAL ends up with a very small <0.15 advantage in game time for a goal E, but also gets through the game faster at RTA.

I will repeat: the game appears to be running fractionally faster RTA in PAL.

I also tested that with a dummy run of just acceleration from the start with no L/R, to the first tree crash, and PAL was faster there too.

Obviously this sort of result requires proper TASers to investigate. It is well above my capacity. I must therefore also question the wisdom behind the quote:

"Remember, if you're driving at 293km/h at 60Hz (the full speed in OutRun SMS), it means you're driving at 83% of the speed at 50Hz and therefore driving at 243km/h, which makes it a lot easier and not competing on the same level as other runners."

Accordingly I immediately recommend a removal of any language that presupposes PAL to be unfair by comparison. I do not have an easy solution to establish historical accuracy of board's IGT v RTA problems, except to suggest runners do mule runs on lag strats, or other testing. I saw Chessjerk's recent streams where RTA was way off, and can't explain it.

I will make a post over at TASvideos to see if anyone is interested in this, and I will also ask CecilMonk if he is able/willing to investigate, as he was instrumental in creating a split tool which edited faulty IGT in Super Hang-On.

I also recommend any speed issues be checked by your communities in Genesis versions, and that's all I will post. Good luck!

Thank you for the info PetaQ! Very interesting stuff. Given what you've provided, I think that a community discussion of timing methodology would be useful going forward. I agree that we should take care not to imply PAL has an "unfair" advantage over NTSC, particularly in light of the tic rate inconsistencies you discovered. In the meantime, I would suggest that agreeing upon timer start/stop frames for RTA should be priority one, so that legacy runs can be retimed.

Thanks for looking in to this @petaQ! Means a lot.

I'm running Outrun 3-D at the moment and there are inconsistencies there as well. Especially on Stage 3, I'm losing 80/100 seemingly for no reason whatsoever. Full speed, similar driving.

In my opinion I think we should keep the IGT as long as possible. It's such an iconic part of the game and makes it differ from all the RTA games out there. It also the main reason I run this game in the first place.

Thanks again :)

I've now added a specific category for 50FPS PAL runs available here: https://www.speedrun.com/outrunsms?h=50hz-runs-goal-a-50hz&x=8244l3w2-rn1y6xdn.1w4824vq

That should prevent any future issues on this topic :) but if anyone has another idea, feel free to give your input, of course!