Early XP

[xdesperado]

fallen Cory Trese
Having played through the escape from Red Hill repeatedly I've noticed a significant difference in XP earned early game before arriving in Oskahold.
You recently boosted the XP for ROUS and Shaman which is nice, but there is still a bit of a balance issue as far as XP earned goes. The biggest difference I'm seeing is because of the spiders. If you get spawns of the Brown spiders they are tougher, harder to kill and so there are usually fewer of them. The Green spiders are easy to kill but come in larger swarms generally. Thing is they both give the same XP and this can have a big impact on the parties XP.
I've had play throughs of this segment where entire party is at 6th level by the time they exit Krellin Deeps and others where no one gets to 6th level until after battle at Oskahold gates.
While long term this isn't as big an issue it does have an impact on how the party advances and when they level which can make certain encounters easier and also affect things like resting either to level or to recover HP/SP.
With a faster early XP gain your likely able to rest a bit less frequently, and combine those rests with leveling the heroes. On a slow early XP gain you may find yourself either needing to use potions to recover SP so you can do another fight or two so you can level or else resting to recover and then having to choose either to rest again after a single encounter or continue without benefit of the new level.

Not a major issue, especially long term but would be nice if this could be slightly better balanced somehow. The number of encounters is fixed, only the random factor of type and numbers is creating this situation.

Love HoS even when it has me totally frustrated with something that's happened.
Please keep up the great work with all your games.

[fallen]

xdesperado - thanks for the post. Our design includes provisions that to be sure that the average number of mobs spawned within any room type has a nice even distribution. If you fight enough spider groups, then you'll end up with very similar XP to someone else who fought a lot of spider rooms, even if the first room you encountered was big and the first room they encountered was small. This same mathematical model is propagated through all the mob groups in the game.

When the levels are very low, and the total number of times you've fought spiders is very low, then you can see a difference. You'll fight more spiders soon and you'll evened out nicely over time!

[Cory Trese]

Very sorry you're frustrated we will keep trying to improve.

We will keep looking at the data from players and reviewing the e-mail and Steam posts from new players about the XP rates!

Thanks for the feedback!

[xdesperado]

Thanks for the responses, and I'm not frustrated by games design but because I'm trying to make a NMI run and getting messed up early lol.

Think you may have miss understood my initial post though.
Party level effects XP earned. Getting early large swarms of whatever sort will level party faster at start. If small encounter size early slows progression then you end up behind the curve slightly so that your earning less XP from later encounters. Fast XP early can put you slightly ahead of curve.
With good early XP and careful management of which heroes get bonus kills you can easily have 1 or more heroes at level 18 before Baron fight at end 9f episode 1, conversely below norm early XP can mean even feeding say Selen/Tamilin bonus kills they are unlikely to break level 18 before Baron and rest of party may barely break level 17 or even still be level 16.
Like I said it's not a major issue, but it can have a noticeable impact on play.

[Cory Trese]

Thanks for the feedback!

I ran 12,000 groups through the first 19 zones today and I'll take a look at the data and we'll see if it is an issue that should be addressed.

[xdesperado]

Mar 31, 2016 16:04:05 GMT -5 Cory Trese said:

Thanks for the feedback!

I ran 12,000 groups through the first 19 zones today and I'll take a look at the data and we'll see if it is an issue that should be addressed.

Lol 12,000! Yeah that should definitely give a very good sample.
Like I said this is low priority, sorry if I gave you extra work.

[fallen]

Mar 31, 2016 14:38:44 GMT -5 xdesperado said:

Party level effects XP earned. Getting early large swarms of whatever sort will level party faster at start. If small encounter size early slows progression then you end up behind the curve slightly so that your earning less XP from later encounters. Fast XP early can put you slightly ahead of curve.
With good early XP and careful management of which heroes get bonus kills you can easily have 1 or more heroes at level 18 before Baron fight at end 9f episode 1, conversely below norm early XP can mean even feeding say Selen/Tamilin bonus kills they are unlikely to break level 18 before Baron and rest of party may barely break level 17 or even still be level 16.
Like I said it's not a major issue, but it can have a noticeable impact on play.

Yes, but you can get the same effect by clearing every monster fight or playing faster and clearing less. You can actually gain a far more significant difference by clearing a single extra fight than you can get by fighting a large mob of spiders or a small mob of spiders.

Also, as mentioned above, we've used a mathematical formula to keep an even distribution of the XP you can get out of any group of monster fights. I understand that when the numbers are really small, it looks significant, but all we can do is deploy math and then its random. And enough random will fall into the statistical model.

The only option to address this would be to (1) remove all randomness from every monster spawn in the game and (2) force every player to fight the exact same fights in the exact same order. It sounds like a very different game

[xdesperado]

Lol ok fallen.
Oh and I'm a completionist so I search out every zingle encounter I can before moving on to next area/map.

[Cory Trese]

I absolutely agree that removing any of the randomization would disappoint and anger our core players.

It looks like the maximum theoretical plus/minus is right around 6.7% with the average delta off the mode being 2.1% and in 12,000 games the two largest achieved deltas were 5.8% and 5.1% ... so that 6.7% probably occurs in about 1 in 40,000 games.

Examining these numbers and comparing them to our metrics on player performance on the same levels confirms that the math is correct. Also went back through all the Steam asks and this has not been mentioned yet, which is further good news.

To see if this type of statistical range is appropriate we compared the results with several other core games on Steam and found that Heroes actually provides a lower than average randomization and advancement curve. Comparisons showed that on average single-player competing titles in our genre tend to have around 10% on the same scope. I also looked at some popular multiplayer games (figuring they'd have even tighter ranges) and found the two most popular examples leave around 8% of Gold/XP to randomization allowing a maximum theoretical swing of around 16% which far exceeds HoS.

Having spent around 3 hours on this issue today I'm happy to report that my curiosity is satisfied and that I need to get back to working on ST2. For me, at least, this is case closed.

Thanks for playing and giving us a chance to do the analysis. I wish we'd found something to change that could address your frustration without costing us other players but there just isn't anything here that will let us win more customers.

[xdesperado]

Cory Trese thanks again for your time and efforts.
Also note this wasn't causing me frustration it was just something I'd noticed and thought I'd mention.
My frustration lies in trying to keep my NMI team alive through at least the Great Shaman.
Mistakes on my part or just bad luck in some cases have so far been doing me in lol.