This is my attempt at trying to compare HL with FoL using maths. If you cba with maths, you can skip to the chapter "Conclusion".
There have been many threads the last weeks regarding HL vs FoL, in which there werent really any people who cared to do some real maths to prove that one spec is superior than the other, so i decided to waste one afternoon of my life to solve what possibly has been the most discussed issue among holy paladins since TBC. Please keep in mind that i'm not biased towards any spec, nor am i influenced by most peoples opinion that FoL sucks. Also please forgive me if some sentences make little or no sense, i'm still working on my english. Some of the calculations done below might be rough or slightly wrong, believe them at your own risk.
I am now going to describe briefly the specs i am comparing.
HL spec: Gemming, gearing, speccing, glyphing
I find this section rather pointless since most of you know what HL spec like but i'll go through it again for the sake of new holy paladins / future holy paladins.
This is the basic HL spec for ICC25, keep in mind there are MANY variations of this, the most common being 51/20/0 if you dont raid with a resto druid. There is also the imp conc aura variation, the imp HoJ variation, the not-imp-bow variation, the 51/0/20 variation which pretty much died with 3.2 but some peeps still use it out of sheer noobishness, etc etc etc
Gemming is 20 int everywhere with a NM on head, or a Purified/Luminous combo for some extra spellpower.
Glyphing is Seal of Wisdom, Beacon of Light, Holy Light.
Enchanting is both abit obvious and of little importance.
Gearing is: 2pc T10 264, offset pieces preferably from ICC10HM with haste+mp5, but ICC25 pieces with haste+crit work aswell.
Trinkets are Solace 258 and Meteorite.
FoL spec: Gemming, gearing, speccing, glyphing
The FoL pally can either use the HL spec mentioned above, OR you can ditch divine sacrifice for a 51/5/15 build that offers you 5% crit versus the raid shieldwall. I'd go with raid shieldwall but there are occasions where i'd rather have 5% crit (to mention a few: having 3 more paladins with it in the raid, or fighting a boss whose damage is rather stable throughout the fight like Rotface). Calculations for FoL spec will be done assuming same talent spec as HL.
Gemming is spellpower everywhere, with the 3% increased crit healing meta.
Glyphing is Seal of Light, Flash of Light, Beacon of Light.
Enchanting is same as HL spec, just SP+crit faction enchants instead of SP+mp5.
Gearing is: 4pc 245 T9 (not everyone has access to 258), ICC25 crit+haste pieces until 676 haste, then crit+mp5. Trinkets are double solace.
Now lets see the difference in stats that those 2 specs have, and how they influence different styles of healing:
Stat differences
SPELLPOWER/THROUGHPUT: Assuming 22 sockets and including PvP libram, 2nd spellpower trinket and SP flask instead of regen flask, FoL should have approximately 1100 more spellpower than HL (about 1300 actually, but i removed the SP granted from Holy Guidance to the HL spec as well as the SP loss from having to use 245 gear over the 264 that HL spec uses), however this only applies to FoL, HS and SS only benefit out of 750 extra spellpower. SoL glyph, FoL meta and FoL glyph increase FoL healing by 5+1+2.5% = 8.5% increased healing for FoL, 6% for HS/SS.
REGEN: Both specs have the same mp5, thus im going to calculate the regen provided from HL spec's int versus the illumination regen provided by the FoL spec's crit over a 5 min fight.
HL spec: approximately 800 more intellect. Assuming 90% replenishment and one and a half plea (sometimes you plea once, sometimes twice, so 1.5 pleas sounds fair) and including starting mana you get roughly 540 mp5. Count in the ilvl 200 libram, the insightful meta and you have 690 mp5. Glyph of SoW, assuming 3 HLs every 5 seconds, makes that 840 mp5.
FoL spec: roughly 300 points worth of haste converted to crit due to soft haste cap. That's 6.7% crit, or 48 mp5.
So HL spec has roughly 800 more mp5 than FoL on a 5-min fight.
Now im going to use these results in realistic scenarios that dont have to do with ICC:
1. Single-target healing, infinite mana
Single target healing: no bacons, no ICC healing done/taken modifiers, no mana limitations.
HL build: Holy Light spam. 1.3 seconds cast time. 3700 SP, 50% crit.
Each HL will land for ((5166 + (3200*1,66)) * 1,12) * 1,05 * 1.25 = 15402 healing per HL, or 11848 HPS.
FL build FoL spam. 1 second cast. 4800 SP, 61.6% crit.
Each FoL will land for ((832 + (4800*1)) * 1,12 * 1,05 * 1,085 * 1.308 = 8951 healing per FoL, with a 1491 tick, for a total of 10442 HPS. Adding the SS absorb value (650*0.75)/6 we have 10523 HPS.
So HL does 1325 increased HPS in this scenario. That means HL is 1.13 times better than FoL in this case. Lets move on to next scenario.
2. Single-target healing, infinite mana, bacon taken into account
Single target healing, no ICC healing done/taken buffs, no mana limitations:
HL = 2*11848 = 23696 HPS
FoL = 10523+8951 =19474 HPS
Thus HL is 1.21 times better than FoL in this case. Moving on.
3. Single target healing, healing target near death
This is a very tricky scenario. We have to assume worst case possible: target at 1 HP and, divide into 3 categories: Target being attacked by boss that does less than 10523 dps, boss that does more than 10523 dps but less than 11848 DPS, and boss that does more than 11848 dps.
Less than 10523 dps: The chance for a FoL to save your target is 100%. The chance for HL to save your target is 1/1.3 = 0.76%, depending on whether you cast it 1.3 seconds before next boss attack or after.
More than 10523 dps, less than 11848 dps: The chance for a FoL to save your target is 0%. The chance for HL to save your target is 0.76%.
More than 11848 dps: Both spells have a 0% chance to save your target.
Realistically, in ICC25HM the 3rd scenario is more likely, but it's a "neutral" scenario (doesnt affect FoL vs HL debate). Now this is very rough calculation but i'd say scenario 1 happens on 20% of the bosses, scenario 2 on 10%, and scenario 3 on 70%.
Thus, the chance FoL will prevent the target from dying in this kind of situation is 1*0.2 + 0*0.1 + 0*0.7 = 0.2 or 20%.
The chance HL will prevent the target from dying in this kind of situation is 0.76*0.2 + 0.76*0.1 + 0*).7 = 0.228 or 22.8%
Therefore, HL is 1.14 times superior to FoL for healing a target near death. Moving on to next scenario:
4. Raid healing - 6 targets + 20% icc buff
Why 6? Because 6 is the holy paladin magic number. HL+splash = 6 targets. Number of people you can keep FoL HoT on = 6. Makes sense now, doesnt it? I didnt want to include icc buff but that would be simply unfair towards HL spec, and as i said im not biased towards either spec.
HL: 15402 + 5 * (15402/10) * 1.05 * 1.2 = 25104 HP every 1.3 second, or 19310 HPS.
FoL: (500 + 0.75*4300)/2 + ((8951*1.2))/2 + ((8951*1.2)*2)*6)/12 = 17974 HPS.
Thus, HL is 7.4% better than FoL at 6-target healing. Moving on.
5. Single-target healing, 5 minutes, not infinite mana
I'm not going to calculate mana costs of spells other than HL and FoL. This is pure HL/FoL spamming. If you want, you can increase the mana useage by calculating the mana cost of bacon/SS but your results wont be different than mine.
HL: Starting mana pool 44k, 800 mp5 more than FoL.
Starting mana pool + mp5 over 5 minutes = 92000 mana
1,5 divine pleas = 16500 mana
Replenishment (90% uptime) = 23760 mana
Meteorite Crystal = roughly 3k mana.
Total = 135.260 mana
Mana useage: 1050 mana every 1.3 seconds => you will last 167 seconds before going oom.
FoL: Starting mana pool = 32000
1.5 divine pleas = 12000 mana
Replenishment (90% uptime) = 19200 mana
Total = 63200 mana
Mana useage: 308 every second => you will last 205 seconds before going oom.
Thus, in a 5-min fight, FoL will be able to heal for 38 more seconds compared to HL. That's 22% more healing time.
The Grand Finale
So what was this all about?
My theory is that every fight in the game can be analysed into those 5 scenarios i mentioned before. My goal is to try and understand which of these 5 scenarios correspond to each ICC25 HM fight, and then combine the results of the scenarios to come up with a final number on each fight. I strongly believe this is the most mathematically strict way to prove which spec is better. Are you ready for it? Lets go then.
Marrowgar is scenarios 2,3,4 and 5.
Thus HL is 1.21 * 1,14 * 1,074 / 1.22 = 1.214 times better than FoL.
LDW is scenarios 1,3 and 5 (and 4, but to a lesser extend, so i wont include it)
Thus HL is 1.13 * 1.14 / 1.22 = 1.05 times better than FoL
Saurfang is 2 and 5.
Thus FoL is 1.22 / 1.21 = 1.008 times better than HL.
Dreamwalker is 1, thus HL is 1.13 times better than FoL (it's actually more if you can bacon her and heal a tenacity pet but w/e)
Sindra is 2,3 and 4, thus HL is massively better than FoL.
Princes is 2,3 and 5.
Thus HL is 1.21 * 1.14 / 1.22 = 1.13 times better than FoL.
Queen is 2,3,4 and 5.
Thus HL is 1.21 * 1.14 * 1.074 / 1.22 = 1.21 times better than FoL.
Rotface is 1, 3 and 5.
Thus HL is 1.13 * 1.14 / 1.22 = 1.06 times better than FoL
Fester is 1,3,4 and 5.
Thus HL is 1.13 * 1.14 * 1.074 / 1.22 = 1.14 times better than FoL
Putri (havent done that yet on HM) is 1,2,3 and 5 (and 4 to a lesser extend, wont include)
Thus HL is 1.13 * 1.21 * 1.14 / 1.22 = 1.27 times better than FoL.
Finally, LK normal is 1,2,3,4 and 5.
Thus HL is 1.13 * 1.21 * 1.14 * 1.074 / 1.22 = 1.37 times better than FoL
Conclusion
To be honest i am a bit surprised with these numbers. Yes they are very rough and yes i was expecting HL to be better than FoL in most fights but i definitely wasnt expecting a clear triumph of HL over FoL. I guess it can be safely said that HL > FoL in 3.3 content.
Thank you for taking the time to read this, constructive criticism is always appreciated.