Difference between revisions of "Experience"
m (→Primary Stats)
|Line 164:||Line 164:|
* Once you have strong Psi soldiers, you can easily use [[Experience Training]] to optimize advancement.
* Once you have strong Psi soldiers, you can easily use [[Experience Training]] to optimize advancement.
*In the endgame (when you have tons of cash), if you're a micromanager
*In the endgame (when you have tons of cash), if you're a micromanager to soldiers maxed across the board, 's one , ones who in a . other stats , , and the to .
== Regarding Caps ==
== Regarding Caps ==
Revision as of 06:30, 28 March 2015
Experience, in X-COM, is the accumulation of combat experience by successfully performing actions. After combat, this experience is applied towards the improvement of stats.
This article discusses how experience is earned and how it affects a soldier's statistics and skills.
How Experience Points Are Applied
Soldier stats can be divided into two areas, primary and secondary stats.
- Primary Stats are stats that are directly influenced by the actions that they are associated with. These actions can be called Primary Actions. They are:
- Every time a Primary Action is taken, Primary Skill Points are earned. They count the number of successful primary actions taken. This is used to calculate how much experience is gained at the end of combat. For specifics on each stat, refer to the individual stat pages by clicking on the links above.
- The number of primary skill points earned at the end of a mission relies solely on the number of primary actions performed during combat, as shown below. Current skill is irrelevant. Whether you have the value that can be rolled for a recruit, or you are one point below the cap.
- Points assigned have a stairstep function, corresponding to a range within which points are rolled. As more actions are performed, the greater the range:
|Ave. Inc. /
|Marginal Inc. / Act. from|
1, 3 and 6 Actions resp.
|1 - 2||0 - 1||0.5||0.50||-|
|3 - 5||1 - 3||2.0||0.67||0.75|
|6 - 10||1 - 4||2.5||0.42||0.40, 0.16|
|11+||2 - 6||4.0||0.36||0.35, 0.25, 0.30|
- Successful Psi actions count as 3 actions. Bravery uses a different approach. See below.
- Performing more than 11 actions won't cause any more increase. Once you reach that figure you get a 2-6 roll and that's that. (Seen in code, and tested to 200+ actions.)
- If there is a limited number of actions your squad can perform (i.e., a limited number of aliens to kill), the "Average Increase / Minimum Actions" shows the most efficient way to go first and the "Marginal Increase" how to proceed from there. The 2.0 average increase for 3 actions is best. In other words, if you let one soldier shoot 3 times instead of letting 3 soldiers shoot one time, the average increase for the squad as a whole is better (here, an average of 2.0 within the squad versus an average increase of 3 x 0.5 = 1.5 with the same three shots). Going from there, aiming for 11 actions provides the best marginal utility (0.25 vs. only 0.16 when going for 6 actions). When there isn't enough to go around, choose who and how you want to advance your soldiers and allocate actions in this order: 3 actions (0.67) and then directly to 11 actions (0.25 per additional action). While you might still want most of your soldiers to take at least 1 action to trigger secondary stat grow (see below), you should never aim for 6-10 bracket as it is better to have 37.5% of you trainees increase +4 (and the others at +2) than having all of them increase +2.5 on avg. for the same number of actions.
- The only time one might have gone for 1 action (besides for triggering secondary stat gain) is if you knew you could only get off 2 more shots, not 3. But short of cheating, Damage is always unpredictable, so instead of taking 1 shot at 2 soldiers when you think your last alien is about dead, you should probably have 1 soldier see if they can get to 3 shots. By the same token, you would prefer to take two soldiers to 6 actions instead of taking one to 9 or 10.
- Notice that any intermediate shots (e.g. your 2nd or 8th shots) are a waste of
alienser, actions unless you reach the next minimum-actions level. You can count bullets as a crude approximation, or use a utility that reads the UNITREF.DAT experience counters from a saved game for exact counts. The number of bullets you've shot will accurately reflect your Reaction counter (unless the person also did direct fire sans reacting), but knowing how many hits were landed for Firing Accuracy can be much more problematic. For a little more detail on efficiency versus experience, see this.
- Psi actions are a special case. Successful panic or mind control (MC) attempts cause the psi experience counter to go up by three actions. Thus, you only need to do 4 successful MCs or panics to get all the psi skill points you're going to (4x3=an experience count of 12). But unsuccessful attempts only cause the counter to go up by one. So you need 11 failed attempts to get to the 2-6 roll range, just like for all other primary stats.
- Throwing Accuracy is also an oddball. While it follows the rules for all primaries (except Bravery), it only increases your Throwing Accuracy. It does not trigger secondaries to increase. (It would've been too easy to increase stats, otherwise! Actually, Psi is just as easy to develop, but it takes time to research psi, and then it needs more combats to reach its cap.)
- Letting your soldiers throw 3 times just before ending a mission maximizes Throwing Accuracy growth versus your personal time spent.
- Performing more than 255 actions of a given type will cause all experience to be lost. This is known to be true for Psi actions, and is undoubtedly true for other actions as well as they are stored in one byte fields. Psi is the only place it's realistically likely. More precisely, experience gets wrapped around, and starts counting again. If you are seriously testing stats, or are challenging yourself by allowing only one psi person, the distance from 255 to 265 (255-10) is only 4.3%, or 3 to 4 MCs on top of 85 successful Psis (28 turns' worth, if they've been successfully MCing since Turn 1). Thus, this may sound scary, but you have nothing to fear unless you've had someone successfully MCing 3 times every turn, for 30 turns - and even then, just let them MC a few more. What's 10 more compared to 255?
- X-COM uses a simple but robust system where all squaddies are expected to participate, and skills increase in proportion to how often they are used.
Bravery experience can also be gained on the battlefield, but using a different algorithm from other primary stats. The bravery stat itself is stored in increments of 10 (10 minimum, 100 maximum, with 110 possible for certain non-human units), and it can only increase by 10 points following each combat (it either gains 10, or nothing).
To gain bravery experience, a unit must have its Morale reduced below 50, where it stands a chance of panicking or going berserk. If the unit successfully resists panicking, it will gain a "bravery experience" point. Each bravery experience point gives the unit an additional ~9% chance (or an eleventh, to be precise) of gaining +10 Bravery at the end of the combat. At 11 or more bravery experience points, the +10 Bravery gain is guaranteed.
Like other primary stats, gaining one or more bravery experience points during a combat triggers secondary stat increases.
- Secondary stats increase when one or more primary actions are performed (except for throwing). Secondary stats are:
- Secondary skill points work quite differently from primary skills: It does not matter how many primary actions are taken; all that matters is that you did one primary action (land a non-IN attack on an enemy, react to something, not panic when Morale is below 50, or use a psi attack) -- your secondaries will increase all that they're gonna. It also doesn't matter if you did more than one kind of primary action.
- The maximum possible increase is highest when the stat is lowest (it can increase the most when you need it the most). The amount that the stat can increase falls steadily as your secondary skills approach their caps, down to an average of 1 point (range 0-2) per combat for each stat. The overall formula is that you gain 0 to 2 points, plus a tenth (rounded down) of the remaining increase possible. (For a graph, see this). Thus, secondary stats which start very low should increase quickly, and will then increase less rapidly as they rise. Rookies with the lowest possible starting stats will have the following as their max possible roll:
- The minimum possible roll is always zero for secondaries. Across the four secondaries, soldiers will show wide variations of luck or lack of luck following each combat, just like flipping coins. Still, the odds are that someone low will rise fast. It's real hard to roll a lot of zeros in a row when the maximum is two or more (and it always is).
- If a primary skill is over its cap, it does still trigger (non-capped) secondaries to increase, by performing it.
Given the rules of how experience works, one can predict how long it will take soldiers to advance their attributes.
Primary skill advances can be based on hypothetical soldiers who always get:
- Low experience: 3 primary-skill experience points leading to a 1-3 roll and thus, always an average of 2 points per combat
- High experience: 11 primary-skill experience points leading to a 2-6 roll and thus, always an average of 4 points per combat
Secondary skills likewise use an average roll for advancing, according to the decreasing-slope function described above.
Because you can select or reject soldiers based on their starting stats, the more important question becomes, "which stats take the longest to advance from recruit maximum to its cap?" In other words, values above recruit max can't be "bought"... so which have the farthest to develop and should be focused on, if you want to maximize your stats?
You can use this as a basis for keeping or rejecting new recruits, or otherwise choosing which skills to develop.
The table below shows the number of combats needed for a soldier to advance from recruit maximum to attribute cap, on average. A more detailed look, including starting from recruit minimum, theoretical min and max advancement times, etc., can be found here and a PDF of the same is here.
Putting It All Together
- While you might focus on the averages stated above, the actual points assigned are entirely random. The range is what you should expect for any one soldier on any one combat mission. This follows the usual X-COM behavior of defining a range and rolling randomly within it.
- Do your best to get every soldier you are interested in developing, to get at least one primary action (not including throwing). This will trigger secondary stat increases. Then try to shoot for the big breakpoints of 3 or 11 primary actions. Doing more than 11 actions does not increase a particular primary stat any more, so give other soldiers the chance to perform actions once your favorite soldiers have maxed out on theirs.
- Any soldier can (and will!) ultimately become a maxed-out superman. (Not counting Psi-wimps with a low psi strength, who will never improve at all.) The worse a secondary stat is, the quicker it will rise. And you can try to target who performs which primary actions. Taking advantage of the breakpoints of 3 and 11 can also help you manage potential skill increases (a.k.a. targets) as effectively as possible. That said,
- Early in the game, don't spend much time or money fretting over stats. Any soldier is liable to come up snakes with their Psi strength. And every soldier can otherwise eventually become a superman.
- Once you have strong Psi soldiers, you can easily use Experience Training to optimize advancement.
- In the endgame (when you have tons of cash), if you're a micromanager and wish to have soldiers maxed across the board, the main focus when screening recruits is the stats that can't be improved - Energy Recharge (which is equal to a soldier's starting TU, divided by 3 and rounded down) and Psi-Strength. So one would screen incoming (psi-locked) rookies for TUs, keep the ones with 60 TUs (who have 20 Energy Recharge), and then put the remaining rookies through a month of Psi-Lab training and discard those with Psi-Strength below some threshold (100 might be too strict given the one-month investment of time in each, but 90 or 95 is attainable given sufficient funding and a large number of Psi-Labs). All other stats are of secondary importance, since they can be improved with experience, and as such screening will only save you time training them while further increasing the number of rookies you need in order for one of them to "pass muster".
Every soldier statistic has a maximum allowed value, as shown in the table below. Your soldier will eventually have a value equal to or slightly greater than the cap. It is allowed to go over the cap the very last time you have a combat while it is still under the cap. For example, the maximum Firing Accuracy is 120. If you go into a combat at 119, do lots of shooting, and get a +5 roll, you will walk away with 124 Firing Accuracy - but it will never increase again.
Primary skills can overflow their cap by 5, and secondary skills can overflow their cap by 1, as shown here:
|Primary Stat||Min Cap||Max Cap|
|Secondary Stat||Min Cap||Max Cap|
- *Psi Skill is an exception. Psi Lab training can continue to add 1-3 points each month after reaching the cap, and will loop the value back down to 0 if it passes 255.
You have a 33% chance of getting capped at Cap+1 for secondary skills, and you have no control over what happens (barring reloading your game and ending combat again). The chance is 33% and not 50% because there are two ways to reach Cap (being at Cap-2 and rolling 2 or being at Cap-1 and rolling 1), but only one way to reach Cap+1 (being at Cap-1 and rolling 2).
For primaries, a soldier that does 11+ actions (for a 2-6 roll) every single combat mission will get Cap+5 in that primary skill 5% of the time, on average. "Every single mission" means, without regard for the fact you're nearing the cap (and playing with the possible roll results). You can increase the odds to as high as 20%, if you use the primary increase roll ranges to inch your way up to Cap-1, then do 11+ actions on your final combat. Or you can have a 0% chance of Cap+5 if you don't pay attention and don't do 11+ actions near the cap!
Experience training outside combat
A Rookie that has gained any experience except throwing, is promoted to Squaddie (also see promotions).