Does someone have some real Gameplay frametime or FPS data (CSV/Excel)?

Hello all,
I was wondering if anyone could provide me with some real video game fps or frametime data that I could use for making a tutorial on explaining percentiles? I see a lot of people calculating or using percentiles incorrectly which is why I think it would make for a great topic, starting with explaining how they are calculated (at least the Excel percentile function way) before explaining why they are great in terms of game framerates and charging for bandwidth at a datacenter.

A related video is this one from Gamers Nexus. I see percentiles commonly being portrayed in this way. Percentiles will be calculated from frame times, but they will use the figure in the context of framerate (which is the inverse of frametime), but will not invert the percentile number.

I do want to help, but I don’t have big game library. I could play Nier;Automata run around some X area or play doom (2016) and do some sections. I also have 3DMark / Time Spy to add into the mix.

With a little research on how to record frametimes, it seems like FRAPS isn’t as accurate as something more expensive like FCAT.
Is that alright with you?

Riva Tuner Statitistics Server, RTSS, SDK 7.0.2 has open source RTSSFrametimePipeSample sample application that can record frame time data for any 3D application, could be something to look into.

I think 1% lows etc are included in the setting for the MSI Afterburner overlay nowadays. Yes, MSI AB 4.4.2 has it implemented. Caveat is that you need to have the benchmark function running for it to record those values, so remember to set a benchmark key.

Then there is AMD Ocat that also can be used to capture frame-time data.

Oh I didn’t know that!

Thanks for the effort, I’ll try out.

Digital Foundry has FCAT tools but I’m not sure if it’s available to the general public. It can take a ProRes file from a SSD recorder and do FCAT on it, since ProRes is perceptually lossless.

PCPer also built their own FCAT tools back in the day.