The lead designers were originally going to talk about this topic at BlizzCon, but it didn’t really match the content of the rest of our “Intro to Pandaria” presentation, and seeing as how we finished our 90-minute slot with 93 seconds remaining, there wouldn’t have been room for it anyway. But several of us did bring up the issue with players and media we talked to, and it even ended up in at least one FAQ, so we figured we’d go ahead and get the information out there. Note that unlike much of what we presented for the upcoming Mists of Pandaria expansion, this is not an announcement. It’s more of a problem we’d like to address, and a couple of ways we potentially might do so. Feedback is certainly appreciated.
Big Number Syndrome
Hey, our stats are growing exponentially. If you look at everything from the Strength on a weapon to the damage being done by a Fireball crit or the amount of health the Morchok boss has, they look downright absurd compared to the numbers for level 60 characters in the original shipping version of World of Warcraft. It’s not exactly a surprise that we were going to end up here, and we knew where we were going every step of the way, yet regardless, here we are.
Fig. 1. Item level vs. character level. Brown = vanilla. Green = BC. Blue = LK. Red = Cat.
The numbers grew so much primarily because we wanted rewards to be compelling. Upgrading from a chestpiece that has 50 Strength into one that has 51 Strength is undeniably a DPS increase for the appropriate user, but it’s not a very exciting reward. Such negligible increases can drive players to do some weird things, such as skipping over tiers of gear or entire levels of content. This is particularly relevant when we’re talking about a new expansion. We don’t want level-85 players to have a reasonable shot at level-90 dungeons and raids (or PvP opponents) just because that content is balanced for gear that isn’t much better than what the level-85 players have.
So we arrived at this point in a logical fashion, and we don’t really think we should have handled things any differently. However, it’s still a weird place to be, and it’s about to get weirder. These aren’t real items, in that we don’t know for sure what the item levels will be in patch 5.3 and patch 6.3 (if only we planned that far ahead!) but they are reasonable guesses, and you can see just how ridiculous the items look.
Fig. 2. A theoretical item from patch 5.3.
Fig. 3. A theoretical item from patch 6.3.
So what do we do about it? There are two general categories of solutions. The first is to make the numbers appear more manageable and the second is to actually change the numbers.
The first solution could include changes like adding commas and the like to large numbers. We could also compress all of those 1000s to Ks and all of those 1,000,000s to Ms, much like we do with boss health today. Internally, we have been calling this the “Mega Damage solution” because instead of your Fireball hitting for 6,000,000 damage, it would hit for 6 MEGA DAMAGE (queue the Arcanite Ripper guitar solo).
Fig. 4. Mega Damage. Name/screenshot not to be taken seriously.
If we can make numbers such as floating combat text and boss health and item stats a little easier to read at a glance, then maybe we can endure numbers increasing exponentially for many digits to come. Now there are some very real computational limitations. PCs just can’t quickly perform math on very large numbers, so we’d have to solve all of those problems as well. Even today, tanks can hit the ten digit threat cap on some encounters.
Item Level Squish
The second solution actually involves compressing item levels, which is why we call it the “item level squish solution.” If we can lower stats on items, then we can lower every other number in the game as well, such as how much damage a Fireball does or how much health a gronn has. If you look at the item level curves, you can see that most of the growth occurs at the maximum character levels for the various expansions. This is because we keep rewarding more and more powerful gear to make the new raid tier and PvP season in an expansion reward significantly better gear than the previous one. However, those huge item level jumps don’t accomplish a lot once the character level has increased again. Very few players notice or care how much of an upgrade the Black Temple loot is over the Serpentshrine Cavern loot when their characters are level 80.
With that in mind, we could go back and compress the big item level increases that occur at level 60, 70, 80 and 85. The Mists of Pandaria gear would still grow exponentially from patch to patch, but the baselines would be a lot lower. Health could go from 150,000 back down to something like 20,000. The big risk of this approach is that players will log into the new expansion and feel nerfed… even if all the other numbers are compressed as well.
In other words, your Fireball will still do the same percentage damage to a player or a creature that it does today, but the number would be smaller. Logically, this seems like it would work, and it does. But it feels weird. When we tried this internally, everyone agreed that it just felt off throwing a spell for hundreds of damage when you are used to it doing thousands of damage.
I came up with an analogy -- even though I know logically that people drive on the left side of the street in the UK (we drive on the right side of the street in the US) and wouldn’t be surprised to see it, it would still feel really disorienting if I was driving in the UK and had to make a right-hand turn.
Fig. 5. Item level vs. character level before and after ‘squish’. Brown = vanilla. Green = BC. Blue = LK. Red = Cat
So Now What?
As I type this today, we haven’t decided on which if either solution we want to try. Maybe we’ll come up with yet another solution. Maybe it’s the kind of thing we can put off for another expansion so that players don’t have to adjust to the new talent system and a drastic item level compression at the same time. Or maybe it’s better just to pull the Band-Aid off fast and fix everything at once. Time will tell. I did, however, want to outline the problem lest any of you believe we don’t think there is a problem. There is. We’re just not sure of the best solution yet. If your answer is that stat budgets don’t have to grow so much in order for players to still want the gear, our experience says otherwise, and thus these proposed solutions exist. Your thoughts on the matter are valuable.
Greg “Ghostcrawler” Street is the lead systems designer for World of Warcraft. The last time he used “Fig. 5” in an article, it related fish predation to estuarine hydrocarbon contamination.
Updates to the patch notes below can be found in red. Last updated Nov. 4, 10:00 AM PDT.
The time for testing out many of the upcoming patch 4.3 features is upon us. In this article you’ll find frequent updates to the official patch 4.3 notes as development continues. You can find a list of the major features we’ve discussed thus far on our official forums.
If you’ll be testing this major content patch, be sure to visit the PTR Discussion forum often to share your thoughts with the community. We also recommend visiting the PTR Bug Report forum to document any issues you come across.
World of Warcraft PTR Patch 4.3.0
Notes for all World of Warcraft patches can be found here.
These notes will be updated through the course of the testing process. Please be aware some changes may not go live until a new PTR build is released."
Table of Contents
Click a link to be taken directly to the corresponding section of the patch notes.
The achievement Tol Barad Veteran now requires 25 victories in Tol Barad, down from 100.
Death Knight pets now properly inherit their master’s crit and spell penetration stats.
Dungeons & Raids
General Bug Fixes
Instead of damage numbers which themselves are completely arbitrary...I mean what does doing 5 damage to someone even mean? How about you just move to percentages. Fireballs crits for 25% of total life. Fireball hits for .02% of total boss life. Then the inflation thing goes away. Maybe it would be a little wierd not to see actual numbers, but I wouldn't mind. The only thing that matters is how much you do versus they guy next to you, and it wouldn't feel like a nerf either because the actual damage numbers wouldn't go down, they would just be different.
Squish it. But not to the point of 20k HP... thats Wotlk Life ._. I imagine Cata life would be 40k and Mists would be 60k. Can i has 60k hp plox?
Remember when we did 2500dps at the end of BC?We do 30,000 dps now, and no one sees that as a big deal.So... what's another 50k dps?
Of course it would be across the board. If you lost 90% of your damage the bosses and mobs would lose 90% of their HP totals and so forth.
But GC, I really would like to keep my ilvl 277 weapons :|
Mega damage?Someone at Blizzard is a fan of Rifts.
18,446,744,073,709,551,615 is the maximum limit if WoW was transitioned into using 64-bit unsigned ints where it's currently using 32-bit ones and has a current unsigned limit of 4,294,967,295.18,000,000,000,000,000,000 is a really big number (Its over 9000! By a lot!). Now, going from 32-bit to 64-bit will logically increase memory bandwidth within the computer, but carries the caveat that modern computers are already 64-bit and therefore have to use masking/shifting operations to properly handle non-native 32-bit integers. As a result, there isn't any significant performance loss/gain on a 64-bit machine if WoW was to be converted.The only problem is that the number I mentioned above has no real *meaning* to a person. That's where the "mega damage" concept comes in to save the day. And for a very long time it would probably work just dandy.But, coming from someone who works as a programmer and deals with lots of stats on a regular basis, if you really want the numbers to actually convey something resembling an answer to your boss, it never ever fails to condense all that information down into percentages. X% of loss, X% of gain. Someone else suggested that instead of damage popping up over the creature, a percentile of how much you did pops up. I consider that idea to be extraordinary. Instead of your fireball hitting PvPGuy for 23,358 damage, it hits him for 8%. Then you turn around and hit UndergearedGuy with the same fireball for 22%. Then you get ported to an instance and hit a boss with the same fireball for 0.2%. Your brain can instantly understand three things... A - whether you did a lot of damage or not, and B - how much target the health has, and finally C - how many times you need to hit the target for it to die. Your brain can intuitively understand all of that simply by hitting targets and seeing a percent. "oh, I'll have to hit PvPGuy a little over 10 times. I'll have to hit UndergearedGuy about 5 times. I'll have to hit the boss a whole crapton of times, but only half as many as when I had sucky gear and only hit him for 0.01%."It makes more sense than doing 71,283,583,928,119,732 damage. It makes more sense than doing 72,283.6 Mega-Damage. Percentages are a very relatable concept. This is my endorsement for that idea, however odd it may seem coming from a world where we currently just see huge damage on huge health-pools.
LVL SQUISH for the win :)1 lvl, Fireball does 5 damage, mob has 15 hp, you 3-shot or 2-shot the mob, like it is now @ lvl 1... Smaller numbers, same damage %....Would feel weird, cuz from the graph you'd be critting 10-15k (currently 30 - 50k) with best Cata gear... Also this would keep end-game epics longer in the play after the launch on new expansion, unlike Shadowmourne getting replaced by new quest blue item...But what I don't understand is, how much will current 1 damage count in new system? If they'd keep it 1, not 0.3 or smth, then it would make mobs harder (a little)And for people wandering if they will be able to one-shot Lich King, then the answer is yes... Damage will be lower, so the HP of everything...Why not just make every number in percentage % ? That would make things a lot easier...
The numbers are smaller obviously =)
I'm all for the Squish... personally I'm hoping they put up a test for it on the PTR well before MoP to see how it feels outside of internal testing and also so they can implement it on live servers WELL before the 5.0 changes
Just a quick question, won't reducing the stats cause some weird hickups with stat priority and/or damage for certain classes.Currently some classes benefit from different stats more at different numbers (making sense, i hope?)If everything gets %based squished it should be largely the same, but you just now they're will be some weird math issues :p
i think we shuld keep going along the same path and not change any thing as far as the numbers problem or power creep goes. and use this as experience and make things nice and smooth for titan.i think players like big numbers, but there is almost a dragon ball Z effect where things just keep getting bigger and bigger until one day im pugging a raid and LFM in trade chat , and i ask my friend to look up some warriors item level on wow-heroes and he comes back tells me that wow-heroes says that his gar score is over nine thousand and then im like what ? over nine thousand ? THATS IMPOSSABLE !
I'll have to say I like the ilvl compress idea. Don't get me wrong, I like have 174k hp unbuffed as a tank, but I did like it back in BC where there was a HUGE difference between the tank with 14k hp and 16k hp. It made people see that every little bit of stats did matter, not huge chunks in cata.So I'll hop onto the ilvl compression. Hopefully that'll get people to start seeing the little things a smidge more often eh?