So given that gear drops are much fewer than in the past, I’ve decided to stop guessing as to which gear will be an upgrade and start simming.
I was talking to someone yesterday and they said anything that uses stat weights like AskMrRobot or Pawn is going to be disadvantageous compared to running a sim which compares your character as it is, against your character with said item equipped.
This true?
I went to Raidbots and simmed my character twice, once for the stat weights and once doing the second method i just mentioned. The stat weight comparison said a shoulder piece I had just recieved, which was a 15 ILVL upgrade with stats I don’t want, was 6% better than my current piece according to the stat weights raidbots gave me. This according to the string it generated for me that i had put into Pawn. I’m not sure where 6% comes from but that’s what it had said. Then i simmed my character using the gear comparison tool and it said that my character as it is currently would do only 1 whole DPS point more right now, than if i were to equip the new shoulders.
Which is more accurate and reliable? Both came from raidbots.
Also yes, I do intend to keep the new shoulders regardless so i can fudge my ilvl a bit for PUGs.
See, THIS is what actually matters. Performance is so variable based on environment, circumstance, and event, that there is no level playing field.
I use Pawn and it’s stats weights.
Why?
Because it’s easy.
The weights are “good enough” at the 5%+ range. At the 1% range, it just doesn’t matter. An extra lucky crit gives me 1%, right?
But pawn is simple. It’s “good enough”, and it always about “good enough” since there can be no best. My spec gimps me enough as it is.
And always look back over your encounters and ask where a couple pts of DPS would have changed the outcome. I bet you find very, very few, if any events as such.
Save for actual DPS checks, mechanics > DPS every time.
I might be wrong but from what I understood from more of my end game friends that main stats are more important thing expansion than your secondary stats. So typically if something is at least 3 ilvl higher it’ll just be better, Until you get around 200 ilvl then you can be more concerned with your secondaries.
True or not, I don’t know xD Decided to stop caring so much for high end raiding lol
Stat weights are meant to give a general idea, not be 100% accurate. Simming your character and picking the piece that gives a boost is also not 100% accurate because the sims emulate perfect robot play - and we are humans that do not play perfectly every millisecond, but they are usually more accurate than stat weights.
every playing field is level, 2 mages fighting the same boss who are the same spec are on the same level playing field minus gear, talents, etc. the only time simming could really be wrong is when it doesn’t take garbage gameplay into account. yes serpentstalker’s trickery is best Hunter legendary, but if a player is so bad they don’t like to use aimed shot a lot and just spam arcane or rapid fire that would probably make Surging Shots better.
that doesn’t mean it’s the best it just means for that player it would be best and lets be honest that type of isn’t going to figure that out either way they’re going to run what icy-veins etc says being obivious to the fact they’re really bad.
Any true test of sims comes from actually engaging in content since there are so many factors at play when you are doing damage - everything from executing perfect rotation, to boss mechanics, to your internet latency. Sims are also known as “Patchwerk DPS”, based on the Naxx raid boss - the only raid boss in the game with no mechanics; you just stand still and DPS until dead. Rarely would you actually do the DPS numbers it says for any other boss.
Nowadays, people tend to lean more towards Raidbots, since it’s based on SimulationCraft or “simc” - the standalone simming program. From what I’ve read some time ago, simc APLs are worked on by more people and is more accessible, so it tends to be a little more accurate (but not perfect) compared to the closed-source AMR.
As Vedimh had said, stat weights give you a general idea and aren’t something set in stone. There may come a point where stats will have diminishing returns or a breakpoint limit and then you’d have to re-weigh your stats to see what to focus on next, hence why stat-weighting isn’t perfect.
Reading the Curseforge page, the stats that come preloaded with Pawn is based on AMR, so I rarely use that data and instead upload the weights from Raidbots into it and use that template instead.
But as I said, actually doing dungeons/raids is the best test. Holding on to your pieces is a good thing to do, so you can play around with different combinations to see what will provide you the best performance.
Raidbots actually uses updated, correct APLs to sim most of its stuff.
AMR is utter trash and only bad or extremely uninformed players use it. They use a private, internally-generated APL that is very, very clearly flawed as hell because it gives suboptimal results 24/7 and the guy that runs it is an utter manchild on social media.
You can adjust sims and the APL to account for multiple situations (i.e. 40 seconds of AoE with no AotD for Unholy), and sims are far less flawed than the extremely antiquated AMR and Pawn.
Better still there are sites that give you decent numbers to create your own pawn stat weights so you have a decent idea if the item is an upgrade. Mostly though as long as you roughly follow the stat priority of a site like wowhead and keep in mind that 10 ilvl pts is usually an upgrade, then you can’t go wrong. The only time you have to question whether something is an upgrade or not is with trinkets and rings. For example, there’s no point in replacing a best stat ring with a high ilvl worst stat ring.
Sure, but I’m perhaps making the gross assumption that the stat weightings “out of the box” in Pawn are “good enough”. If they have the base priorities roughly in order, that’s fine. The nuances are less important in the big picture.
I’ll never forget Wrath. One aspect I didn’t like was I was constantly getting gear and I really had no idea whether any was “better” or not, and we didn’t even have publicly available iLevel to go on.
So, my bag filled with a lot of redundant gear until I got sick of it and sold it all away, but kept in “indecision” until then.
Pawn settles that for me, while I have enough fundamental understanding when things get close.
Trinkets, of course, are the worst as nothing evaluates well the proc or on use affects, followed by rings. But ready insight in to the stat priorities can help with that.
And besides, for me, haste is always a win. Anything to speed up my zillion hard castes is better FOR ME, who cares what the meters say.
You stat weights change depending on a lot of factors. When you get to the higher levels of gearing you should sim to find the best stat weights for you at that point of your character development. Then you put those weights into a mod like Pawn so it can use them to find you better gear.
If you just go with the default weights then you’ll get advice based on what fits a certain set of people but not the best advice for your exact situation.
Go back to raidbots, to the part where you simmed your stat weights. There should be a big warning sign, telling you that stat weights have inherent disadvantages and direct gear sims are better. If you click this warning, it will take you to an article explaining why in exhaustive detail.
Why would you trust these people to compute your stat weights, but not trust them when they tell you there’s an issue with stat weights?
