It might not be you.

spinnytop · June 2017

So we just did a little experiment on PTS. Two of us made our characters almost completely numerically identical stat-wise, and used the same power rotations. One character did significantly more damage than the other, an amount large enough to be startling. This makes me wonder if there's some sort of bug causing some people to do significantly less damage than other people even if they have the same stats and are using the same powers - this is CO after all, so it's not out of the question.

So just be aware that it might not be your fault if your dps isn't as good as some other people's.

baelogventure · June 2017

That sounds like shenanigans.

ealford1985 · June 2017

Well....there is a thing called a @crosschan Luck, so it's possible.

darqaura2 · June 2017

The hell???

blockwave · June 2017

Latency, the time it takes each players to do the rotation.

nepht · June 2017

Cryptic Maths.

darqaura2 · June 2017

On a serious note the amount of effects, latency, server issues and other "out of the user's control" affecting dps ability between two people with the same exact build . . . wouldn't be surprising in this game.

nepht · June 2017

darqaura2 said:
On a serious note the amount of effects, latency, server issues and other "out of the user's control" affecting dps ability between two people with the same exact build . . . wouldn't be surprising in this game.

So what are you saying here? If your closer to the server you will perform better? I have never once considered that. Wonder if a proxy would effect that. Would explain why I can always beat a buddy who uses the same build as me but they live in a country that requires them to use a proxy.

spinnytop · June 2017

I would understand latency making one person do ~2k more dps if one of us had terrible latency, but that wasn't the case. Also, this was done in the powerhouse, so there wouldn't be tons of effects occuring.

Latency is, of course, another factor on its own, but I don't think it's the primary culprit with this rather large discrepancy.

darqaura2 · June 2017

> @spinnytop said:
> I would understand latency making one person do ~2k more dps if one of us had terrible latency, but that wasn't the case. Also, this was done in the powerhouse, so there wouldn't be tons of effects occuring.
>
> Latency is, of course, another factor on its own, but I don't think it's the primary culprit with this rather large discrepancy.

Well . . . crap

darqaura2 · June 2017

pantagruel01 · June 2017

Last time I saw a dps discrepancy that large, one person was in hybrid role.

magpieuk2014 · June 2017

It's not that PTS bug where you create a new character at level 6, give it some levelling gear, then level it up to 40 and find that the gear hasn't levelled up as well? You generally have to unequip something before it re-calculates the stats.

theravenforce · June 2017

This is interesting...

Are you able to give more specifics as to what led to this and how others can test?

Also, any difference between you and the other player?

spinnytop · June 2017

magpieuk2014 said:
It's not that PTS bug where you create a new character at level 6, give it some levelling gear, then level it up to 40 and find that the gear hasn't levelled up as well? You generally have to unequip something before it re-calculates the stats.

nope, both characters were copies of lvl 40s that hadn't had their levels changed.

roughbearmattach · June 2017

The next test:

One player copies the same character twice.
Then test the DPS on each copy.

This could help determine if issues are character-based or account-based.

After that, have another player log on as you, but in a different location, with the copies of characters you just tested.

This could help determine if the issues are location/ISP-based.

kallethen · June 2017

Do we have some hard numbers on this?

sandukutupu · June 2017

spinnytop said:
So we just did a little experiment on PTS. ...
So just be aware that it might not be your fault if your dps isn't as good as some other people's.

Hello I work on the Neverwinter wiki a little bit. I added some info there about DPS Parsers. You can read more about that here on the wiki. Now before everyone gets into a EULA discussion on 3rd party software, let me state that a parser is just something that reads your combat log and can be used either with the game active or inactive. If Cryptic starts banning you for using parsers, they could ban you for using Notepad as well.

I do recommend the second one I mention on that page. "CombatLogParser" is hardwired for Cryptic games (no plugin needed) ACT will need a plugin to decode the combat log. The combat log is activated through chat and records all combat DPS from all players in your area.

blackwingsoul · June 2017

@sandukutupu

Thanks for posting this! I was just asking about an alternative to ACT. I can't test right now, but I'm hoping CLP will be a little more consistent.

opalflame · June 2017

I've noticed that the numbers on my attacks and heals will increase or decrease randomly sometimes. This is with the same number of stars, same gear, same build, same role, same number of form stacks and everything. Maybe this has something to do with it?

kallethen · June 2017

I'm still gonna take this with a grain of salt as it currently this is just an anecdotal observation.

pantagruel01 · June 2017

The other thing to be aware of is that how you run your parse can significantly affect dps values. I generally assume that parses from two different people are not directly comparable.

cptmassive1 · June 2017

What about skill? Not to bash any player. Rotations work on timing and some are better than others and latency plays into that. I have had variations in my dps (not counting RNG shenanigans) up to ~1000 just by being either fat fingered or smooth on my activation timing.

When I am out traveling I have to relearn the timing for the network that I am on. I know my home network plays into my success as I am lucky to have better internet than everyone else. It spoils me for other places.

spinnytop · June 2017

cptmassive1 said:
What about skill?

The rotations we used were very simple, it was vs target dummies so there was nothing to block, and the tests were done for a good duration. We basically removed skill as a factor.

pantagruel01 said:
The other thing to be aware of is that how you run your parse can significantly affect dps values. I generally assume that parses from two different people are not directly comparable.

Can you provide specific details for this? If people are gimping themselves with their parser settings and reading 2k less dps than they are actually doing then that is valuable information that people can use. Thanks.

kallethen said:
I'm still gonna take this with a grain of salt as it currently this is just an anecdotal observation.

That's fine, if it doesn't apply to you then you don't need to worry about it.

kamokami · June 2017

This seems very odd. If you really want to get to the bottom of it then I'd suggest redoing the test with some more controls in place.

1. Simplify the rotation down to just spamming the main attack. Make sure you are both fully charging it or if it's a maintain then you're both fully maintaining it. No AOs, debuffs, etc......just literally pressing one button over and over. You can introduce more complexity into the rotation one power at a time later.

2. Make sure you both start logging only when you both have 8 form stacks to eliminate any potential difference in how fast each of you got there.

3. Make sure you both start and stop logging at the same time. Set up a keybind that will say something like "started" and "stopped" in local chat so that the other person knows to also start/stop logging on their end.

4. Log for at least 3-5 minutes to account for any user error.....incomplete charges, etc...just to build in some "practice" time.

5. Both of you should log each other and yourselves....so just start attacking the test dummies at the same time and start logging at the same time. By the end you should have 4 DPS numbers to compare:
- Foxi's Parser: Foxi DPS #1 vs. Friend DPS #1
- Friend's Parser: Foxi DPS #2 vs. Friend DPS #2

You can then check by subtracting Foxi DPS from Friend DPS in both parses and seeing if the difference is the same. You can also check to see how closely Foxi DPS matches in Foxi's Parser vs the Friend's Parser. This way you can account for any differences in parsing methods or user error with regards to starting/stopping the combatlog.

spinnytop · June 2017

1, 2, 4, and 5 were already in place.

pantagruel01 · June 2017

spinnytop said:
Can you provide specific details for this? If people are gimping themselves with their parser settings and reading 2k less dps than they are actually doing then that is valuable information that people can use. Thanks.

It's usually about start and end time of parse; to get the most consistent results, pick a specific condition under which you will turn parsing on (for high sustained dps, probably reaching 8 stacks of your form) and off, and use a fairly long parse period.

kamokami · June 2017

spinnytop said:
1, 2, 4, and 5 were already in place.

So what were the four DPS numbers? (from #5)

spinnytop · June 2017

kamokami said:

So what were the four DPS numbers? (from #5)

~5k, ~3k, ~3k, ~5k

I'll mention it again, we're talking about a difference of ~2k dps. The things people are mentioning might lead to a small discrepency, less than 1k. That's not what we're talking about here. This isn't a case of "oh I did a few hundred less".

It might not be you.

Comments