XP in Beta3 is still badly skewed

[KGB]

Piranha,

I see that the XP rewards in Beta3 are still badly skewed.

It's turn 10 in our game. You get a hero in the city right next to me. Attack my hero and win and then jump to L4. All thanks to 1 battle in 1 turn.

I'll say the same thing I did in Beta2. That's not good. A hero arriving next to you happens sometimes. It is what it is.

But skewing the rewards from battles is not good. It's very bad for the game. My own hero that was killed was my initial turn 1 hero. In the preceding 10 turns he searched 1 ruin and conquered 3 cities (2 neutral and one of yours with 2-3 men in it). All that and he still wasn't even L2.

There is FAR too much XP handed out for battles. FAR too much. To get to L4 you must have gotten 4-5K worth of XP.

There are 2 things that should be done both of which are trivial to do:

1) The battle report said you were 80% likely to win thanks to your Dreadlord + Demon combo. You should not get full XP from any battle you are more than 50% likely to win. The XP reward should be based on the challenge of the battle. You divide the losers percentage chance to win by yours and multiply by the XP. So a 50% chance battle gives 50/50*XP=full XP. As a 25% chance (you are the underdog) gives 75/25=3x XP (3 times the amount). An 80% chance (what you had) gives 20/80=.25 XP (1/4 the amount). If the loser had 0% chance to win then give a minimal amount (maybe 25 or 50xp).
2) Limit the amount of XP gained in 1 battle to be 1 level worth (your next hero level whatever that is). So that even if you get 10K from a battle you'd be limited to what you need to reach the next level (1K in the case of L2, if you were L5 and needed 5K to reach L6 then you'd get 5K). This is the model that Dungeons and Dragons uses to keep things from getting out of control when you fight boss battles.

KGB

[joasoze]

+1
Strongly agree to this

[piranha]

The XP calculations is just like that.

My total STR: 49
Your total STR: 43

49/43 = 1,1395

43 / 1,1395 = 37,7

37,7 * 104 = 3920 XP

If we should limit the XP to max one level per fight that will also affect other things. For example the learning skill may not be very useful. The paladin heros 140% learning wont be veru useful.

Fighting big battles is usually a big risk, your risking a whole lot more than when you fight a bunch of neutrals or go into a ruin. I went from level 1 to level 4, so 3 levels and that was with a level 1 hero. I don't really se the problem. My hero is not a killer machine now.
In fact it's not enough to get a single point of command. So my hero is the exact same as before.
In beta2 this would have meant +1 str, +2 move, +1 command. In beta3 it will get you some move or invidual strength but not a single point of command. I don't think its getting out of hand. After playing a bit I think it will turn out that getting a killer hero will be quite hard.

I think its worth testing for a while before jumping to conclusions, if its too easy to level we should see heroes that are killer machines but I'm still waiting to see it.

[KGB]

Piranha,

If the strengths were 49-43 then something is wrong with the combat calculator. It said you were 80% likely to win. 80% which seems extremely strange given the strengths were that close. So the idea of high risk/high reward doesn't apply here as I don't consider 4-1 favorites being a risky battle.

Anyway the XP should be based on winning percentage, not strength differential since that's what really determines the difficulty of the battle.

Limiting to 1 level won't affect the Paladins learn skill adversely. That's because later levels take more XP to reach so there will be fewer times you get capped by the level limit. Plus the learning skill will still be useful on neutrals, smaller fights and one assumes quest XP in upcoming versions. I strongly think jumping multiple levels from 1 fight is a bad idea.

As far as a killer hero goes, you are +2 when you attack thanks to the Dreadlord bonus. You can at L4 take another +1 stack attack so you are in effect +3 stack when attacking. That's *HUGE* at turn 10. It makes Heavy Infantry 6 strength. Had you gotten an ally like a Devil you'd be +4 combat now with the ally bonus. That is the very definition of a killer stack when players have <30 men in total.

KGB

P.S. One of the issues with the XP model you've chosen is that all the reward is based on big risky battles. There is so little XP awarded from ruin searching, neutral capturing and winning small battles (2-3 units) that it takes 10+ such battles to get to L2. Then another 10-15 to get to L3 and more to get to L4 etc. So its exceptionally hard to get your initial heroes up to decent levels compared to someone who happens to have a hero join in just the right spot to be able to fight several large battles. Something seems inherently unfair about that (staying alive a long time, moving over a large amount map area) given all prior Warlords games worked on the reverse model (steady XP flow from lots of battles/quest/city captures/ruin searches). Maybe you can just award all heroes 100XP a turn for every turn for survival/training so that living for 20 turns would get you to L3 even if you did nothing. That doesn't seem game breaking while at the same time it helps initial heroes who don't get lots of XP early in the game.

[piranha]

If I got the devil (which is harder to get) I would have the same since the dread hero have a stack bonus.

At the end I had a demon which was STR 10 and units facing it would lose 1 str due to negating the city wall. So I would not face 43 str since the demon will kill a couple of units and lower your total STR on each simulation.

I have started work on a javascript simulator and if it will give me other results then I know the simulator is wrong, but I did quite a lot of testing and controlled quite a lot of tests.

[KGB]

Piranha,

If I got the devil (which is harder to get) I would have the same since the dread hero have a stack bonus.

I didn't know that it was a stack bonus. It says it's a +1 attack bonus which I thought was like the Elementals +3 attack bonus which is totally separate from a stack bonus (Pegasi for example) as those definitely stack together. So you are saying it's not like the Elementals bonus then? Now that's confusing as I would never have guessed that adding a Pegasi to a Dreadlord made no difference to the bonus's.

Does this also mean then that a Devil would cancel such a bonus even though it comes from a hero?

At the end I had a demon which was STR 10 and units facing it would lose 1 str due to negating the city wall. So I would not face 43 str since the demon will kill a couple of units and lower your total STR on each simulation.

I have started work on a javascript simulator and if it will give me other results then I know the simulator is wrong, but I did quite a lot of testing and controlled quite a lot of tests.

Actually I believe the simulator when it says 80% since I know how much time we both spent running results in beta2 and we were getting the same numbers with 2 entirely different programs.

My point wasn't I thought the 80% is wrong. Rather that simply using 49/43 for the XP ratio is wrong when the actual combat difficulty is 80/20.

KGB

[piranha]

There are basiclly two types. Individual bonuses and stack bonuses. Stack bonuses that are based on terrain/attack/defend/antiair, and individual bonuses based on terrain/attack/defend/antiair.

Stack based bonuses always count as stack based bonuses regardless if they come from terrain or from attack. So the devil will cancel that bonus.

It might be better to go by that percentage. I think its too early to start changing things, I want to see more results first.

[KGB]

Piranha,

There are basiclly two types. Individual bonuses and stack bonuses. Stack bonuses that are based on terrain/attack/defend/antiair, and individual bonuses based on terrain/attack/defend/antiair.

I see. Definitely wasn't clear until just now. So the extra +1 the Dreadlord gets is basically pointless if you already have a Pegasi. So really once you are past the initial few turns and start having production going that bonus is almost pointless.

Regardless of how you generate the XP formula I still maintain allowing heroes to jump multiple levels from 1 battle isn't a good thing. I hated it in Beta2 and I still hate it in Beta3 even when my own heroes do it. It just feels cheap. 1 battle, 1 level (at most).

My analogy is that my first hero is a guy who joins a company, spends 10 years (turns) working on several projects (cities/ruin conquering) that are quite valuable for my company (empire). For that he has received no promotion despite being integral to their success. Your hero is the bosses son who joins the company and is immediately put on this great team that's near the end of a big project (you gathering armies). He basically rides their coattails for a short time and then gets promoted from junior employee to CEO because the project is a success.

Another analogy is that my first hero is the guy who works hard and saves money his whole life hoping to be a millionaire one day. He's on his way but isn't there yet. Your hero is the guy next door who just turned 18 bought his first lottery ticket and hit the jackpot for multiple millions.

KGB

[piranha]

I have no problem with the young guy winning millions on the lottery as long as it me

You could also say that being part of small skirmishes in some half defended villages is not going to give you the same experience as leading a huge army vs a well protected city controlled by the hero who has 10 years of experience .

I understand what you mean. There will some time to watch beta3 games now, I'm more worried about the weird combat result you got.

About the bonuses, thats right and thats probably why I dont consider that bonus as unbalanced. Sure its nice to get a "free" pegasi but movement and vision with the amazon is really nice to, atleast I found the extra vision has been very useful.
Not sure how I could make it easier to understand all bonuses. I tried to explain them in the warpedia but I could probably explain the hero bonuses in more detail.

[KGB]

Piranha,

I have no problem with the young guy winning millions on the lottery as long as it me

Me either!

You could also say that being part of small skirmishes in some half defended villages is not going to give you the same experience as leading a huge army vs a well protected city controlled by the hero who has 10 years of experience . I understand what you mean.

I wouldn't mind so much if the XP from other things was worth more. I mean the point of the game is to conquer cities, not slay enemy units. So conquering cities should be giving more XP. Plus that battle wasn't a heroic battle between 2 super stacks full of Dragons/Devils etc. It was just scouts, orcs, hv infantry and giants. If you had been a Paladin you'd have gotten 40% more XP (6K worth) and if it had been a real battle between super stacks it would have been 10-12K worth. That's the equivalent of conquering a hundred of neutrals.

Otherwise I get to the point where my hero could conquer 3 enemy cities with 1 man in them or kill 8 hv infantry in the open and he's going to ignore the cites to get the gobs of XP. That makes no sense. Plus it also means that if I can disband men and I see an enemy hero stack I can't beat I'm going to disband rather than attack because I can't afford to give out all that free XP to the enemy. Again that doesn't make much sense.

It would help a lot if conquering a city gave a base 500XP plus units inside and ruins gave 500 for L1, 750 for L2 and 1000 for L3. Then taking a 2-3 neutrals + a 1-2 ruins gets you to L3. That helps offset the gobs of XP from battles.

There will some time to watch beta3 games now, I'm more worried about the weird combat result you got.

Me too! I blame that result for my entire turn of rotten luck where I lost 18 men to kill only 4 despite being a 70+% favorite in all 3 battles! So I resigned because clearly the Gods weren't on my side in that game.

About the bonuses, thats right and thats probably why I dont consider that bonus as unbalanced. Sure its nice to get a "free" pegasi but movement and vision with the amazon is really nice to, atleast I found the extra vision has been very useful.

I agree 100%. Now that I know how it works it's a LOT less useful. Mid to late game it's entirely pointless as you'll always have that kind of bonus available someplace. It's really only useful on small maps where you get climatic battles in the first 10-15 turns.

The Valkrie is a great hero in 1-1 games. You don't have to worry about facing down the enemy in the early turns so you have time to be a ruin whore and she is perfect for that.

Not sure how I could make it easier to understand all bonuses. I tried to explain them in the warpedia but I could probably explain the hero bonuses in more detail.

Just took a look at the new Warpedia. Impressive. You've spent a lot of hours updating it and adding graphics.

Your bonus explanation is fine. It makes sense to me. I guess what's not clear is when you are looking at the hero menu itself and seeing that bonus and realizing it's a stack bonus and not a special attack mode like the Elemental gets.

I guess what would be handy is to list all 3 heroes in the hero section and show exactly what each one starts with. Then you could point out the stack bonus's area and mention it again to help clarify.

KGB

P.S. I looked at the combat section. It says high risk = high reward. Actually it really should say large battles = large reward. I mean if my hero fights a giant 1-1 in a L1 city that is a 50/50 battle. That's every bit as high risk as an 8v8 battle that is 50/50. But I don't get the same high reward and I probably get less of a reward than a strong hero stack does from slaughtering 8 Lt Infantry