Here's the article that I'm looking to rewrite: http://www.smogon.com/dp/articles/ev_distribution edit as of 7th march - I know I've left this sitting around for a while, but I really intend to write this in around a weeks time when the holidays start. What I'm going to do: Remove all the formulae and calculations, or at least remove most of them and present what's left in a more accessible manner. (I'll still be giving lots of examples though!) Give better advice – make the reader more aware of the big picture, while explaining in better detail the intricacies of defensive EV spreads. The guide that this might replace is currently a fourth gen guide. It's possible that we could not replace it but just add this one instead. I'm not quite sure what to do here actually - the guide could apply to gens 3, 4, and 5 but not 1 and 2. Introduction Basic stuff – what evs do. Link to Ev training guide. Rather than learning formulae and calculating, best way to calculate stats from spreads is just to use a resource such as a damage calculator or the teambuilder of a simulator. (link to Smogon Damage calc, Honko calc, PO, PS) EVs and Natures Explanation of how you can waste EVs by using the wrong nature. Boost the higher stat where possible. Speed The Speed stat is the stat where one point can make all the difference, but where many points can equally make no difference. Therefore for this stat precision can be very important Generally, Speed EVs should be invested with something in mind, unless you are using maximum or minimum Speed Look for specific benchmarks Link to Speed tiers articles Think carefully about the pokemon you want to outrun. How common is it? How likely is it to be running this amount of speed evs? Does one pokemon win regardless of speed? How well does my team deal with this pokemon? Comment on analyses and Speed creep – it's often sensible to add that little bit more Speed than the "standard" (4 evs is probably not enough!) Explain why we can't do that in our analyses Quick warning – it might be tempting to add more and more speed evs, jumping from benchmark to benchmark, thinking "what do those 20 HP EVs do anyway?" but they add up. Any number of defensive or offensive evs can make the difference between life and death even if you can't think of any specific examples. The other stats are different to Speed in this respect. Attacking Stats Exceptionally simple, usually Benchmarks are much less critical than speed ones (similar to defensive benchmarks) Splitting evs between stats here is a bad idea unless absolutely necessary - there's no attacking equivalent of HP! Don't go mixed without a good reason. Mixed attackers are extremely useful for breaking through walls though (they tend to be dedicated to one defensive side). But how should you split your attacking evs? 50:50? Well, usually not - generally, a mixed attacker tends to focus on one side, running a staple, STAB move on it, and then will have one or more coverage options on the other side, dedicated to removing would be counters. Therefore, you want to invest the minimum required to take out said counters, and then boost your staple move as much as possible. Offensive example Pokemon (a mixed attacker, maybe MixMence or Hydreigon in OU?) Defenses By far the most complicated aspect of EVing Often you'll want to focus on one defensive stat in order to completely wall physical or special attackers. However, there's no avoiding taking the occasional attack from the wrong side. However, in contrast to offensive stats, the HP stat means that mixed defensive spreads do not adversely affect each individual stat too much – investing in HP usually gives more overall "defensiveness." Investing EVs in lower stats gives larger improvement as a percentage (eg. investing in defense on a Chansey) Benchmarks such as surviving certain moves can be relevant, but there is no need for one. There need not be a specific example for a certain degree of defensive investment. Remember that the damage done by an attacker is somewhat randomised (from 85-100% of the maximum damage) and so there isn't a fixed cutoff as with speed. Efficiency Because there are three variables here, it is possible to waste defensive potential with the wrong spread. It is possible for one spread to give better defenses than another both physically AND specially. If so, the first spread is considered to be inefficient. The quickest way to gauge defensiveness is simply to multiply HP and the given defensive stat. Efficient spreads can be given by X-acts applet (link) Generally, unless the pokemon in question has enormous HP (Blissey, Chansey, Wobb), you should be maximising HP if you are looking to invest in both defensive stats. If one of the defensive stats actually exceeds the HP stat, then you should invest more in HP if possible, even if you are only interested in one stat. If no stats are maximised, but all are invested in, you should consider carefully whether the spread is efficient (it probably isn't!) Tweaking the HP stat Tweaking the HP stat of a set can occasionally prove useful, but this is often overdone at the expense of defensive ability. With the exception of rounding, the method of multiplying the stats is completely accurate. Since damage is always rounded down, rounding of attack damage tends to favour the lower HP spread EVER so slightly. Lower HP is also preferable in case of Leech Seed, Giga Drain etc. However, these things are very rarely relevant – and besides, high HP is good for Seismic Toss, Night Shade etc. Rounding of various forms of residual damage and recovery, is often more important. As the amount of damage/recovery is also rounded down, recovery can be maximised or residual damage minimised by the right spread. Stressing again though that the difference is TINY. E.g if HP is divisible by 16, leftovers will restore 1/16 of your health, whereas otherwise it will heal less (as a proportion). Unfortunately it equivalently maximises sandstorm damage. If HP is divisible by 4, then 3 layers of Spikes, or Stealth Rock for a 2x weak pokemon, will do maximum damage as a percentage of HP. You might have noticed that these are largely at odds, since anything divisible by 16 is also divisible by 2, 4, and 8. If HP is divisible by 16 with a remainder of 1, then this is often called a "Leftovers number" (sometimes also with remainder 0, which is confusing). The idea behind this is that the pokemon in question will faint one turn later than a pokemon with HP divisible by 16 if subjected ONLY to residual damage dealing in 16ths. Ultimately, this is absurdly unlikely unless the pokemon is 4x weak to Stealth Rock. Leftovers numbers are pointless in all but the rarest scenarios. E.g Scizor – this Pokemon is NOT using leftovers, and has maximum HP divisible by 8. It will take one less point of damage from SR and Spikes if you lower the evs in HP to 248. This will pay off after one switchin – in other words a good idea! However, if its HP were divisible by 16 and it used leftovers, this would be a bad idea, because it would immediately regain the lost HP and would gain more if no hazards were up. Defensive Example Pokemon Conclusion