Joined: Nov 23, 2002
Location: on the golf course, in the garden, reading, traveling, and now Consulting
Tue Jan 15, 2013 2:33 am
For your reading pleasure, a most recent Consortium article, written for DM93 Noblish Island and newbies everywhere ..........
TRAINING: SKILLS or STATS
A CONSORTIUM OPINION
EITHER IS A SPIFFY CHOICE
BE SURE TO TRAIN SOMETHING!
Often one hears a statement similar to this: “Train skills; do not train stats.” The implication or sometimes direct statement is that you will be sorry if you train stats, especially certain stats.
The Consortium opinion is that, in the majority of circumstances, it is perfectly all right to train stats, and in some cases it is better to train stats. (This process of training stats is often called “burning”) The purpose of this article is to explore stat and skill training and to and to give the reader some ideas as to if, or when, stat training is either aok or best. The article will also explore when it is considered best not to train stats.
First, to set the stage, there are several principles, givens, and ideas that the reader must understand.
1. Skill raining is mostly dependent on WT. (e.g. The higher the wit, the more skills per fight the warrior is likely to learn.
2. Stat training is mostly dependent on two things:
a. WL; a warrior’s stat learn probability is 5% times the will. (E.g. a warrior with 10 WL has a (5%x10=50%) 50/50 chance to learn the stat, while a warrior with 17 WL has an 85% chance of learning the stat.
b. The number of stats of that type already trained. A warrior’s learning rate for a specific stat is halved each time he earns one of those stats. For example, a warrior with 10 will has a 50% (5%x10) of earning a first stat of a type, but only a 25% (5%x10 halved) of earning the second stat of that type. ((and on to 12.5% for the third)
3. Training most stats also earns the warrior certain skills.
4. There are limits to the number of stats and skills a warrior can have.
a. In basic, a warrior can have no higher that 21 of any stat type.
b. A warrior is limited to earning 20 skills (beyond the initial skills the warrior is “born” with) of each type in his career. Those 20 skills may come from learning skills or from training a stat which gives the warrior one or more of those skills. He is deemed “maxed” when he learns the 20th.
c. A warrior can, however, learn more that 20 skills of a type if he first learns his 20 allowed skills, and then trains stats which give the warrior one of those skills. (This factor is why one often hears “Do not train stats or you will be sorry”.)
5. Certain stats improve the “physicals” of your warrior. (e.g. there is a reasonable possibility that training ST can improve how much damage your warrior does when he hits)
With those principles in mind, one can postulate certain scenarios which might provide a basis for decision-making as to “what to train”. These represent, first, the knowledge necessary, then, The Consortium scenarios for training decision – STATs or SKILLs?
1 Be aware of the probable skill learn rate. (This is a highly variable number, based on many, many factors of the game, including luck rolls.)
2 Be aware of the stat training probability for that stat
3 Be aware that every thing in D2 has a probable roll or luck factor
4 Be aware of the potential “physical” improvement to your warrior for stat training.
5 Know which stat trains do not provide any skills. (E.g. Raising to 12ST improves the stat, but cannot provide any skills or “burning”, while raising to 15ST improves that stat and provides two “burned” skills – attack and parry.)
6 Determine if that specific warrior will be enhanced by a possibility of earning more than 20 skills of a type. (Example, a warrior that is meant for the long-term, or is deemed to be a near-godling is extremely likely to benefit from the extra skills at high-ADM classes where every warrior is maxed or near maxed in stats and skills and every single skill can make a difference in a fight. A warrior who is expected to be around for a shorter period, or is to be “retired” upon graduation to Lord Protectorship, will not be around long enough to earn and benefit from the few extra skills available.
7 Know which is likely to benefit your warrior most when you decide to train skills of a certain stat. (e.g. How many skills am I likely to learn if I train skills verses the probability of learning that specific stat train and the subsequent physical and skill gain earned.)
CONSORTIUM DECISION MODEL
1. Upon rollup, evaluate each warrior and classify – long-term/godling (GOD), possible long term (UNS), standard warrior – run to graduation.(NRM)
2. Do not train stats on GOD, unless they are non-burning stats.
3. Limit training stats on UNS, unless they are non-burning stats, until it is decided that this warrior is not a GOD.
4. If a NRM, train whatever the manager believes will most enhance the warrior winning fights.
5. When deciding between stat or skill training, compare the probability quantity of skill earning vs. the probable of learning the chosen stat and its potential physical and skill contribution.
Perhaps a couple of Consortium examples are in order. Let’s consider these two warriors” 14-7-12-15-20-14-11 striker and 4-10-11-17-17-4-21 aimed-blow.
Although a solid warrior, the striker would be considered NRM, especially since strikers are not considered likely Primus-level styles. The warrior probably rolled good damage and good endurance. The odds of getting a first stat raise are near 100% (5%x20, but everything has a “luck” roll). The warrior at 15WT should expect to earn nearly two skills per turn (Consortium rates 15WT skill learn rate at 1.75ish) Consortium Managers would plan two immediate “bumps” – WL to 21 and ST to 15. Moving to 21 WL gains seven fine skills (2 each Att, Par, Def and 1 Decise) and enhances physicals a tiny bit. (Adds 0.4 it points and improves endurance a small amount) Training skills would have yielded a probability of two skills, and three if lucky. The choice to train WL is a no-brainer. The 2nd plan train to 15ST is also a good one. At 15 ST, two skills are earned, attack and parry. The 15ST also enhances weapon selection, making the striker now strong enough to carry certain “larger weapons” with no penalty. (The battle axe is an often favored weapon of Consortiumites – and probably should be a favorite of yours.) The 15ST also has the possibility of raising the warrior’s damage rating from good to great. Consortium places this damage bump odds at 4-1, appx 20% likely.) Compared to the probability of learning 2 skills (or an unlucky 1 or a lucky 3) the 15ST raise seems like a good one. There are no more planned raises for this warrior, but that does not preclude that a Consortium manager might not later attempt an additional ST raise to 16 (if great damage was not attained) and a raise to 8 CN, especially if the warrior has a “frail” rating.
Warrior #2, the aimer is rated a GOD. There would be no planned stat raises to the warrior, training skill all the way until maxed. However, there would be no reason not to consider CN raises, as no CN raise burns skills and every aimer can use additional hit points. This choice would need to be balanced against the probability of earning 2.25 skills per turn (Consortium rating). One additional consideration should be given. If this warrior rolled “little damage”, (normal damage is more likely) the bump to 5ST could be what is needed to get normal damage. The raising of that stat “burns” a parry and an attack skill, however. For a long-term warrior, especially an aimer which can use little-damage somewhat effectively, this bump would not be recommended.
In conclusion, there is no absolute when it comes to training skills or stats. Training stats is not a no-no. Burning skills can be inconsequential. Physical and skill gains from stats can greatly enhance the warrior and the W/L record. But, like most other decisions in D2, there is a trade-off. Burning stats on a long-term, ADM+ warrior could prove damaging to the warrior’s long-term health. Choose wisely rather than automatically.
Ye Old Consortium Scribe & Crapmaster 2013, Crapgiver 2014
1000 ADM graduates (with 40+ manager IDs) including 105K+ fights and 68K+ wins plus 4 teams with 1000+ wins (Animal Farm DM11 @1518*; Bulldogs DM11 @ 1460; Lenpros DM30 @ 1424; Fandils @ 1336