The Mechanics of Combat SP (Part I)
Overview
The recent early February XIV patch drastically changed the SP system to make it easier to level at all ranks. Since it was implemented, I feel that a number of players have made another attempt at the game, mainly those initially frustrated by the difficult leveling system previously implemented. As one of the major problems with the game, this change will probably go a long way towards improving the population of the game and regaining community trust in the game.
But how exactly does this system work? And more importantly, how can the player base maximize their SP gain?
In this post, I'll go over the basic calculations involving in EXP and SP gain as well as how these rough formulas were deduced from basic observation. I'll include both solo as well as party play. I'll then make a couple discussion points about how to maximize EXP gain under the new system. I've provided some clean tables at the end of the post that will be useful in the future for those interested in predicting their own SP gains. For those not interested in the number crunching and just want the results, I suggest skipping down to the summary section.
The Mechanics of Combat SP (Part II)
________________________________________________________________________________________________
________________________________________________________________________________________________
Initial Solo Data Collection
My initial hypothesis was that the SP system in XIV would closely mimic that the EXP system in XI. The system in XI was based on level difference (or dLVL), where a static amount of EXP was granted for killing a mob of certain dLVL. In this system, killing a 50 mob at 48 would grant the same amount of EXP as killing a 52 mob at 50 (dLVL=2).
I started by trying to find a mob with a very wide, but known level range. I ended up choosing Fat Dodos near Cedarwood since there is a nice 25-34 level range, a lot of them, and they are fairly easy to kill at 26 Archer. The idea here is to just kill these solo until you potentially see 10 different EXP/SP values, with the assumption that each level of Fat Dodo will give a different value for the same level Archer. Once all 10 different values were collected, you could deduce that the highest EXP/SP value corresponded to the highest level Fat Dodo (34), the second highest 33, and so on and so forth. This procedure was repeated for a number of mobs using various jobs not in fatigue (26 Archer, 33 Gladiator, 50 Thaumaturge).
A summary of the initial data I collected is below, including:
- The amount of EXP/SP gained at each dLVL for the mob in question
- The job and rank used to collect the data (not in fatigue)
- The "color" of the mob that was killed
Based on this initial data set, we can see a couple of trends:
(1) SP for fighting different mobs at different levels can be the same as long as dLVL is the same
(2) The amount of SP gained is consistently 40% of the amount of EXP gained
(3) The "con" (color) system is NOT consistent with simply dLVL
(4) There were 2 distinct patterns of EXP/SP gain, which I separated into "Class 1" and "Class 2"
It became clear that the amount of SP gained is simply ~40% of the amount of EXP gained. For the rest of the discussion in this section, I will only refer to the EXP amount gained unless specified otherwise. One can simply convert to SP by multiplying the EXP value by 0.4.
This data set strongly suggested a system similar to XI where dLVL plays a critical role in determining the amount of EXP/SP gained. The 2 separate "classes", or patterns, made this system a bit more complex, however. To further analyze this, I separated these 2 classes, then divided each value in each class by the amount of EXP gained at dLVL=0. This means dividing all values in "Class 1" by 158 and all values in "Class 2" by 188. A table showing this is below:
When we do this, we find that both "Classes" have similar values once corrected with this division. I called the results of this correction the "modifier". The results of this strongly suggest that there are 2 main determinants to calculating EXP gain solo:
(1) The dLVL modifier, which ranges from 0.00 to 3.00. This modifier is arbitrarily set to 1.00 at dLVL=0.
(2) The Base EXP/SP, which is the amount of EXP/SP gained solo killing the mob at dLVL=0.
[ Solo EXP ] = [ Base EXP ] x [ dLVL Modifier ]
____________________________________________________________________________
Analysis - the dLVL Modifier
As stated earlier, I arbitrarily set the dLVL modifier to 1.00 at dLVL=0. To clean up the results into a nicer table...
A couple of points to note from this system that I found interesting. First of all, the amount of EXP you gain is capped after fighting something 10 levels above you. This means that there is no incentive for fighting something over 10 levels above you. The scaling system is also very good when fighting mobs above you (0.20 increases each dLVL); whereas it is pretty slow scaling downward.
____________________________________________________________________________
Data Collection and Analysis - Base EXP/SP
Base EXP/SP is the amount of EXP/SP the mob would give if killed solo at dLVL=0.
While the dLVL modifier system was fairly simple to figure out, what will make this EXP system complex is the varying amounts of base EXP granted for different mobs. In the 7 different mobs tried in the initial data set, there were 2 different base EXP values (called "Class 1" and Class 2"); "Class 1" had a base of 158 while "Class 2" had a base of 188. At this point, I sought out various types of mobs in the game and tried to back calculate their Base EXPs.
This procedure was done most easily by simply soloing something 10 or more levels above me. Because the amount of EXP will be capped at 3x the base value I'm looking for, I can simply take the amount of EXP given by the mob and divide it by 3 to find the base value. Here's an example of this calculation:
This was a Sundrake in the 60-64 level range, making it above dLVL of 10 and capped. Because we know that the EXP is capped, we can just take the 1500/600 EXP/SP value and divide by 3 to find the base values we're looking for. This results in a base value of 500/200. I did similar calculations to this (trying to find mobs I capped on for simplicity) and placed these results in another table. I tried to separate each different Base amount into its own "Class".
The highest modifier I've been able to find was the Sundrake in the screenshot taken above. It is hard to tell exactly how many different base EXP/SPs there are at this point and my testing on this particular aspect of EXP/SP has been limited. There are a couple of possibilities to predict the bases that I've thrown around though:
- Higher level mobs, in general, will have higher bases
- Each species/race will have its own base
- Worst case, every mob has its own programmed base, and would have to be manually tested.
Overall, while this aspect of EXP gain is far from complete, we can draw one major conclusion already:
Mobs with higher bases make more much better EXP targets. Because there is no real incentive to fight things above dLVL=10, the ideal situation is to fight mobs exactly dLVL=10 and with the highest base.
________________________________________________________________________________________________
________________________________________________________________________________________________
Party Based EXP Hypothesis
In XI, the party EXP system was based on the highest level value in the party. This meant that if you took a party of 2 with a level 50 character and level 10 character and killed a level 25 mob, BOTH players would essentially receive nothing (because the EXP is calculated for both players based on the high of 50). Essentially, the EXP calculation for all players in the party was determined off of the highest in the party.
I hypothesized that the system in XIV was very different that XI's system, and that in XIV, the amount gained is completely calculated on an individual player basis. To elaborate, this means that if we take that same situation of a party of 2 with a level 50 and a level 10 killing a level 25 mob, the 50 would still receive nothing, but the level 10 would receive a good amount of EXP. In this situation, the calculation for each player is based on his own individual level, not the high. The reasoning behind this hypothesis is just the observation that you can easily power level characters in the 10-20s on a level 50 character. This sort of situation was impossible in XI.
I generalized how this formula would work to there being 3 components to determining your party EXP/SP:
(1) The amount of EXP/SP one would have gotten had they killed the mob solo.
(2) The total number of players in the party.
(3) The total number of players in the party within the -10/+5 range of each individual.
[ Final EXP ] = [ Solo EXP ] x [ Party Modifier ]
Where the [ Party Modifier ] is a function of the total number of players in party and the number of players within the -10/+5 rank range.
____________________________________________________________________________
Party Based Data Collection
The data collection process for this is fairly simple - you just need to know the EXP you would have gotten had you killed the mob solo, then the EXP you actually did get along with how many were in party and how many were within the specified -10/+5 rank range. I ran a total of 4 data sets for this with the help of my LS "Dancing Mad"
From here, I combined all the party modifiers from all 4 data sets into 1 table...
At this point it gets a bit tricky. The goal is to convert these seemingly random decimals into fractions of whole numbers, where hopefully a pattern can be found. I started by saying that for the values in bold red (where the number of members in range equals the number of total members), there seems to be a pretty strong pattern of a 0.05 drop for every member added. I used this as the basis for forming my fractions. Through some random trial and error, I was able to discern this pattern of fractions...
Basically, the denominators are the total number of members times 10. The numerators were based on the whole number value needed to create the 0.05 modifier drop off when total members in range equals total members. When we do this, the pattern becomes fairly obvious - all columns have the same numerator, while all rows have the denominator of number of total members times 10.
____________________________________________________________________________
Party EXP Analysis
The above table works out quite well in that it seems to perfectly fit every piece of data that was collected. However, one issue I noticed looking at the extrapolated table is that if we continue this trend all the way to complete the table, you'll notice that the modifiers will actually start to decrease as you add more players within the proper -10/+5 range (for instance, look at how a 15 member party scales as you add more players into the -10/+5 range). Without actually recording some data in those ranges, it will be hard to definitively make a statement of accuracy knowing how flawed that system would be.
I will say though that there are some flaws in this system that would lead me to believe that this table still could be correct, even at high numbers of members. If we look at our initial data collection set 1 (on power-level Giant Crabs), the amount of EXP actually goes up as you add players. For instance, having 2 players in party (1 is the power-leveler at 50, 1 is the 'leech') gives you 331 SP (0.50 modifier). Adding another 'leech' in similar rank ups the EXP to 441. Adding a 3rd 'leech' ups the EXP again to 446. Add a 4th 'leech' and it's 449. The system here is a bit flawed in that you will actually get more EXP per kill as you add the number of 'leechers' all the way until 5 total. Given that SE made the oversight here, it would not necessarily surprise me to see them make another one.
________________________________________________________________________________________________
________________________________________________________________________________________________
Results Summary
To briefly summarize the final results and limitations of this...
(1) The amount of SP gained is simply 40% of the amount of EXP gained.
(2) The amount of EXP gained can be found with the basic formula...
[ Final EXP ] = [ Base EXP ] x [ dLVL Modifier ] x [ Party Modifier ]
(3) The [ dLVL Modifier ] is a function of the rank difference between the player and mob being fought.
The dLVL modifier is arbitrarily chosen to be 1.00 at dLVL=0. The amount of EXP gained is capped above dLVL=10 (with a modifier of x3.00). At dLVL<-19, the modifier is 0.00, meaning you will receive no EXP.
(4) The [ Base EXP ] is defined as the EXP gained killing the mob solo if dLVL were 0.
Each kind of mob has its own Base EXP. This Base EXP will vary between mob type and races. In general, higher level types of mobs seem to have higher bases, but no definitive trends have been found so far. The lowest Base EXP in my data set is 158/63, while the highest found so far is 500/200. The Base EXP will have a significant impact on the amount of EXP gained.
(5) The [ Party Modifier ] is a function of the total number in party and the number within -10/+5 of the player.
The table below summarizes the party modifier for all possible situations. But keep in mind that the grayed areas are untested and unverified at this time. I am personally not that confident in the grayed values because it implies that for certain situations, having players out of the -10/+5 range will actually increase your EXP gain (don't trust the gray numbers in this table!). More testing is needed. An alternative table providing exact party modifiers instead of 2 decimal estimates is given inside the post.
>>> THIS TABLE IS OUTDATED DUE TO THE 1.15b PATCH <<<
>>> THIS TABLE IS OUTDATED DUE TO THE 1.15b PATCH <<<
(6) Things still unknown and shortcomings
- Rounding error associated with the formulas will only predict EXP within roughly +/-2 points
- The party modifier table needs some verification at higher number of total members
- The base EXPs of various mobs and races need to be found (hopefully a trend is found)
- The mechanics of Leve linking and Guardian's Favor need to be discussed
- The mob "con" (color) system is still poorly understood
At this point, there are some limitations to the results. For instance, much of the party modifier table is extrapolated from a set of maybe 10 data points (there just happened to be a nice pattern). Also, the calculations in general just have some rounding error associated with it. Since these formulas are based on observation only and not code, I cannot tell you the precise values; however, I do feel confident in these formulas correctly predicting EXP/SP gain within +/-2.
In the future, I hope to first verify or shore up the party modifier table as well as work a bit more on the base EXP values to see if there's any sort of pattern involved. It would be a huge chore to attempt to find the base EXP values for every single mob in the game individually. In addition, I hope to get more into how one can use knowledge of the system to maximize SP gain in both Leve groups and even parties. While there are some shortcomings to these results, I do think that the overall concepts of EXP/SP gain and how to use the system to one's advantage can be discerned from this.
Til next time!
________________________________________________________________________________________________
________________________________________________________________________________________________
The recent early February XIV patch drastically changed the SP system to make it easier to level at all ranks. Since it was implemented, I feel that a number of players have made another attempt at the game, mainly those initially frustrated by the difficult leveling system previously implemented. As one of the major problems with the game, this change will probably go a long way towards improving the population of the game and regaining community trust in the game.
But how exactly does this system work? And more importantly, how can the player base maximize their SP gain?
In this post, I'll go over the basic calculations involving in EXP and SP gain as well as how these rough formulas were deduced from basic observation. I'll include both solo as well as party play. I'll then make a couple discussion points about how to maximize EXP gain under the new system. I've provided some clean tables at the end of the post that will be useful in the future for those interested in predicting their own SP gains. For those not interested in the number crunching and just want the results, I suggest skipping down to the summary section.
The Mechanics of Combat SP (Part II)
________________________________________________________________________________________________
________________________________________________________________________________________________
Initial Solo Data Collection
My initial hypothesis was that the SP system in XIV would closely mimic that the EXP system in XI. The system in XI was based on level difference (or dLVL), where a static amount of EXP was granted for killing a mob of certain dLVL. In this system, killing a 50 mob at 48 would grant the same amount of EXP as killing a 52 mob at 50 (dLVL=2).
I started by trying to find a mob with a very wide, but known level range. I ended up choosing Fat Dodos near Cedarwood since there is a nice 25-34 level range, a lot of them, and they are fairly easy to kill at 26 Archer. The idea here is to just kill these solo until you potentially see 10 different EXP/SP values, with the assumption that each level of Fat Dodo will give a different value for the same level Archer. Once all 10 different values were collected, you could deduce that the highest EXP/SP value corresponded to the highest level Fat Dodo (34), the second highest 33, and so on and so forth. This procedure was repeated for a number of mobs using various jobs not in fatigue (26 Archer, 33 Gladiator, 50 Thaumaturge).
A summary of the initial data I collected is below, including:
- The amount of EXP/SP gained at each dLVL for the mob in question
- The job and rank used to collect the data (not in fatigue)
- The "color" of the mob that was killed
Based on this initial data set, we can see a couple of trends:
(1) SP for fighting different mobs at different levels can be the same as long as dLVL is the same
(2) The amount of SP gained is consistently 40% of the amount of EXP gained
(3) The "con" (color) system is NOT consistent with simply dLVL
(4) There were 2 distinct patterns of EXP/SP gain, which I separated into "Class 1" and "Class 2"
It became clear that the amount of SP gained is simply ~40% of the amount of EXP gained. For the rest of the discussion in this section, I will only refer to the EXP amount gained unless specified otherwise. One can simply convert to SP by multiplying the EXP value by 0.4.
This data set strongly suggested a system similar to XI where dLVL plays a critical role in determining the amount of EXP/SP gained. The 2 separate "classes", or patterns, made this system a bit more complex, however. To further analyze this, I separated these 2 classes, then divided each value in each class by the amount of EXP gained at dLVL=0. This means dividing all values in "Class 1" by 158 and all values in "Class 2" by 188. A table showing this is below:
When we do this, we find that both "Classes" have similar values once corrected with this division. I called the results of this correction the "modifier". The results of this strongly suggest that there are 2 main determinants to calculating EXP gain solo:
(1) The dLVL modifier, which ranges from 0.00 to 3.00. This modifier is arbitrarily set to 1.00 at dLVL=0.
(2) The Base EXP/SP, which is the amount of EXP/SP gained solo killing the mob at dLVL=0.
[ Solo EXP ] = [ Base EXP ] x [ dLVL Modifier ]
____________________________________________________________________________
Analysis - the dLVL Modifier
As stated earlier, I arbitrarily set the dLVL modifier to 1.00 at dLVL=0. To clean up the results into a nicer table...
A couple of points to note from this system that I found interesting. First of all, the amount of EXP you gain is capped after fighting something 10 levels above you. This means that there is no incentive for fighting something over 10 levels above you. The scaling system is also very good when fighting mobs above you (0.20 increases each dLVL); whereas it is pretty slow scaling downward.
____________________________________________________________________________
Data Collection and Analysis - Base EXP/SP
Base EXP/SP is the amount of EXP/SP the mob would give if killed solo at dLVL=0.
While the dLVL modifier system was fairly simple to figure out, what will make this EXP system complex is the varying amounts of base EXP granted for different mobs. In the 7 different mobs tried in the initial data set, there were 2 different base EXP values (called "Class 1" and Class 2"); "Class 1" had a base of 158 while "Class 2" had a base of 188. At this point, I sought out various types of mobs in the game and tried to back calculate their Base EXPs.
This procedure was done most easily by simply soloing something 10 or more levels above me. Because the amount of EXP will be capped at 3x the base value I'm looking for, I can simply take the amount of EXP given by the mob and divide it by 3 to find the base value. Here's an example of this calculation:
This was a Sundrake in the 60-64 level range, making it above dLVL of 10 and capped. Because we know that the EXP is capped, we can just take the 1500/600 EXP/SP value and divide by 3 to find the base values we're looking for. This results in a base value of 500/200. I did similar calculations to this (trying to find mobs I capped on for simplicity) and placed these results in another table. I tried to separate each different Base amount into its own "Class".
The highest modifier I've been able to find was the Sundrake in the screenshot taken above. It is hard to tell exactly how many different base EXP/SPs there are at this point and my testing on this particular aspect of EXP/SP has been limited. There are a couple of possibilities to predict the bases that I've thrown around though:
- Higher level mobs, in general, will have higher bases
- Each species/race will have its own base
- Worst case, every mob has its own programmed base, and would have to be manually tested.
Overall, while this aspect of EXP gain is far from complete, we can draw one major conclusion already:
Mobs with higher bases make more much better EXP targets. Because there is no real incentive to fight things above dLVL=10, the ideal situation is to fight mobs exactly dLVL=10 and with the highest base.
________________________________________________________________________________________________
________________________________________________________________________________________________
Party Based EXP Hypothesis
In XI, the party EXP system was based on the highest level value in the party. This meant that if you took a party of 2 with a level 50 character and level 10 character and killed a level 25 mob, BOTH players would essentially receive nothing (because the EXP is calculated for both players based on the high of 50). Essentially, the EXP calculation for all players in the party was determined off of the highest in the party.
I hypothesized that the system in XIV was very different that XI's system, and that in XIV, the amount gained is completely calculated on an individual player basis. To elaborate, this means that if we take that same situation of a party of 2 with a level 50 and a level 10 killing a level 25 mob, the 50 would still receive nothing, but the level 10 would receive a good amount of EXP. In this situation, the calculation for each player is based on his own individual level, not the high. The reasoning behind this hypothesis is just the observation that you can easily power level characters in the 10-20s on a level 50 character. This sort of situation was impossible in XI.
I generalized how this formula would work to there being 3 components to determining your party EXP/SP:
(1) The amount of EXP/SP one would have gotten had they killed the mob solo.
(2) The total number of players in the party.
(3) The total number of players in the party within the -10/+5 range of each individual.
[ Final EXP ] = [ Solo EXP ] x [ Party Modifier ]
Where the [ Party Modifier ] is a function of the total number of players in party and the number of players within the -10/+5 rank range.
____________________________________________________________________________
Party Based Data Collection
The data collection process for this is fairly simple - you just need to know the EXP you would have gotten had you killed the mob solo, then the EXP you actually did get along with how many were in party and how many were within the specified -10/+5 rank range. I ran a total of 4 data sets for this with the help of my LS "Dancing Mad"
From here, I combined all the party modifiers from all 4 data sets into 1 table...
At this point it gets a bit tricky. The goal is to convert these seemingly random decimals into fractions of whole numbers, where hopefully a pattern can be found. I started by saying that for the values in bold red (where the number of members in range equals the number of total members), there seems to be a pretty strong pattern of a 0.05 drop for every member added. I used this as the basis for forming my fractions. Through some random trial and error, I was able to discern this pattern of fractions...
Basically, the denominators are the total number of members times 10. The numerators were based on the whole number value needed to create the 0.05 modifier drop off when total members in range equals total members. When we do this, the pattern becomes fairly obvious - all columns have the same numerator, while all rows have the denominator of number of total members times 10.
____________________________________________________________________________
Party EXP Analysis
The above table works out quite well in that it seems to perfectly fit every piece of data that was collected. However, one issue I noticed looking at the extrapolated table is that if we continue this trend all the way to complete the table, you'll notice that the modifiers will actually start to decrease as you add more players within the proper -10/+5 range (for instance, look at how a 15 member party scales as you add more players into the -10/+5 range). Without actually recording some data in those ranges, it will be hard to definitively make a statement of accuracy knowing how flawed that system would be.
I will say though that there are some flaws in this system that would lead me to believe that this table still could be correct, even at high numbers of members. If we look at our initial data collection set 1 (on power-level Giant Crabs), the amount of EXP actually goes up as you add players. For instance, having 2 players in party (1 is the power-leveler at 50, 1 is the 'leech') gives you 331 SP (0.50 modifier). Adding another 'leech' in similar rank ups the EXP to 441. Adding a 3rd 'leech' ups the EXP again to 446. Add a 4th 'leech' and it's 449. The system here is a bit flawed in that you will actually get more EXP per kill as you add the number of 'leechers' all the way until 5 total. Given that SE made the oversight here, it would not necessarily surprise me to see them make another one.
________________________________________________________________________________________________
________________________________________________________________________________________________
Results Summary
To briefly summarize the final results and limitations of this...
(1) The amount of SP gained is simply 40% of the amount of EXP gained.
(2) The amount of EXP gained can be found with the basic formula...
[ Final EXP ] = [ Base EXP ] x [ dLVL Modifier ] x [ Party Modifier ]
(3) The [ dLVL Modifier ] is a function of the rank difference between the player and mob being fought.
The dLVL modifier is arbitrarily chosen to be 1.00 at dLVL=0. The amount of EXP gained is capped above dLVL=10 (with a modifier of x3.00). At dLVL<-19, the modifier is 0.00, meaning you will receive no EXP.
(4) The [ Base EXP ] is defined as the EXP gained killing the mob solo if dLVL were 0.
Each kind of mob has its own Base EXP. This Base EXP will vary between mob type and races. In general, higher level types of mobs seem to have higher bases, but no definitive trends have been found so far. The lowest Base EXP in my data set is 158/63, while the highest found so far is 500/200. The Base EXP will have a significant impact on the amount of EXP gained.
(5) The [ Party Modifier ] is a function of the total number in party and the number within -10/+5 of the player.
The table below summarizes the party modifier for all possible situations. But keep in mind that the grayed areas are untested and unverified at this time. I am personally not that confident in the grayed values because it implies that for certain situations, having players out of the -10/+5 range will actually increase your EXP gain (don't trust the gray numbers in this table!). More testing is needed. An alternative table providing exact party modifiers instead of 2 decimal estimates is given inside the post.
>>> THIS TABLE IS OUTDATED DUE TO THE 1.15b PATCH <<<
>>> THIS TABLE IS OUTDATED DUE TO THE 1.15b PATCH <<<
(6) Things still unknown and shortcomings
- Rounding error associated with the formulas will only predict EXP within roughly +/-2 points
- The party modifier table needs some verification at higher number of total members
- The base EXPs of various mobs and races need to be found (hopefully a trend is found)
- The mechanics of Leve linking and Guardian's Favor need to be discussed
- The mob "con" (color) system is still poorly understood
At this point, there are some limitations to the results. For instance, much of the party modifier table is extrapolated from a set of maybe 10 data points (there just happened to be a nice pattern). Also, the calculations in general just have some rounding error associated with it. Since these formulas are based on observation only and not code, I cannot tell you the precise values; however, I do feel confident in these formulas correctly predicting EXP/SP gain within +/-2.
In the future, I hope to first verify or shore up the party modifier table as well as work a bit more on the base EXP values to see if there's any sort of pattern involved. It would be a huge chore to attempt to find the base EXP values for every single mob in the game individually. In addition, I hope to get more into how one can use knowledge of the system to maximize SP gain in both Leve groups and even parties. While there are some shortcomings to these results, I do think that the overall concepts of EXP/SP gain and how to use the system to one's advantage can be discerned from this.
Til next time!
________________________________________________________________________________________________
________________________________________________________________________________________________