Request for participants

[damufunman]

Hypertrophy calc is off...
Didn't square the .6 (not much of a difference) but you also added an extra 15 reps with the 15*.6 (I'm seeing hypertrophy day as 5x3 then 4x4 at 70%, no backoffs at 60%...) which changes it to just 31*(.7)^2 = 15.19, I would say more reasonable, though possibly still a bit high.

In any case, I was looking at this as more of a more appropriate volume equivalent than just total reps or even straight tonnage, not as a complete picture.

Oh, whoops, yeah. Edited to fix that. I still get about 20. So still off by about a factor of 2.

The 15 reps are close grip bench AMRAP.

Oh forgot about the AMRAP set... though for benching I think the hypertrophy work would be more fatiguing for me than 3x5 @80%... though I could be remembering wrong. My week immediately preceding Hanley's programming for bench was:

H: 160, 170, 175, 165 x 8, 150 x 9 (AMRAP) [64%, 68%, 70%, 66%, 60%] normalized tonnage: 4405
S: 185, 200, 200, 195, 195 x 5 [74% (last warmup), 80%, 80%, 78%, 78%] normalized tonnage (incl. wu): 3805 (w/o wu): 3121

Equivalent e1RM with 5x3 + 4x4 @70% and 9 AMRAP at 60% gives NT of 4607... maybe the 3x5 on strength day is not enough?

[PatrickDB]

@damufunman Yeah. I think 5x3 for lower body may be enough for strength day, but I'm wondering if they pressing can tolerate more.

[Hanley]

In the meantime, for peeps on a 1-week cycle, I think the arbitrary 5# increase to e1rm is the best load progression strategy (maybe 2.5# increase for upper).

For us I'm wondering if it's best to just do the arbitrary increases (5 for lower, 2.5 for upper/wk) and just test at the end of the 4 week block via AMRAP (to 9.5/10 RPE) at 80%ish.

Yeah. For the folks who have the full 4-weeks blocked in, they'll simply see "TEST". I'm thinking of leaving the choice of test (1rm or amrap) up to the lifter (which is total bullshit from a "research perspective" but the more lifter-friendly way to do things, I think.

[cgeorg]

Relating to the discussion above, I haven't done any reading or research on this, but how does an intensity-weighted tonnage sound to you guys?
[equation]IWT = weight \times reps \times \%1RM \quad (1)[/equation]
Which can also be represented as
[equation]IWT = 1RM \times reps \times \%1RM^2 \quad (2)[/equation]
This factors in the fact that lower intensities are less stressful for a given number of reps. For example, why 5 sets of 5 with just the bar as warmup isn't meaningful volume. Equation (2) has a [math]\%1RM^2[/math] term, that effectively gives more weight to higher intensities, which I think follows with what people experience of higher intensities are more fatiguing.
Note that this does ignore how the volume is distributed, ie 5 x 3 @ 80% vs 3 x 5 @ 80%.

Normalizing so 1 RM = 1, let's assess the bench hypertrophy and strength days using this idea.

Hypertrophy: 31*(.7)^2 + 15*(.6)^2 = 20.59

Strength: 15*.8^2 = 9.6.

Not looking good.

I need to look up what Nuckols thinks the shape of the regression curve should be for this. Then I could use Hanley's hypertrophy and strength day values to get a formula.

ETA: fixed calculation

Reps per set needs to factor exponentially I think, as does %1RM.

[equation]sets*(reps)^x*(\%RM)^y[/equation]

Find values of x and y that make sense based on experimental data. Let's try x=1.2, y=2.

Hypertrophy: 70%, 3x5 then 4x4
[math]3*6.9*.49 + 4*5.3*.49 = 20.5[/math]

Strength: 5x3 at 80%
[math]5*3.7*.64 = 11.8[/math]

Nah, I think rep count to a factor of intensity might actually be a better metric. Or (1-intensity) to a factor of rep count.

[equation]sets*(reps)^(\%RM)[/equation]

Hypertrophy: 70%, 3x5 then 4x4
[math]3*5^.7 + 4*4^.7 = 19.8[/math]

Strength: 5x3 at 80%
[math]5*3^.8 = 12[/math]

This is probably approaching correctness. I don't think it will scale linearly with set count though. Hmm. Lets square root the sets.

Hypertrophy: 70%, 3x5 then 4x4
[math]3^.5*5^.7 + 4^.5*4^.7 = 10.6[/math]

Strength: 5x3 at 80%
[math]5^.5*3^.8 = 5.4[/math]

Intensity to a factor of reps? No, normalizing this to sane values is dumb. Eh it's probably just something like (bar speed as %1RM barspeed) * %1RM of weight used. Could also be that 5x3@80 really isn't as fatiguing. If you jack y up enough in the early equation you could make it work. Maybe x should be less than 1? Ok one more try.

[equation]sets^.5*reps^.5*\%RM^2[/equation]

Hypertrophy: 70%, 3x5 then 4x4
[math]3^.5*5^.5*.7^2 + 4^.5*4^.5*.7^2 = 3.9[/math]

Strength: 5x3 at 80%
[math]5^.5*3^.5*.8^2 = 2.5[/math]

These seem close. Ugh I lied I thought of something else, but it would require converting reps @ percents into RPE. Something like

[equation]sets^x*(RPE/10)^y*\%1RM^z[/equation]

Solve for x, y, z.

[cgeorg]

^ my guess is that x is something like .4, y is 2, and z is very near 1. Ugh, RPE chart doesn't go that low. x6@70 is @6, so we'll call x5@5 and x4@4. x3@80 is @6.

Hypertrophy: 70%, 3x5 then 4x4
[math]3^.4*.5^2*.7 + 4^.4*.4^2*.7 = .47[/math] <- I like the scale of that number to describe that workout

Strength: 5x3 at 80%
[math]5^.4*.6^2*.8 = .55[/math] <- Hey I found an equation to fit our assumptions. Done.

[SpinyNorman]

Yeah. For the folks who have the full 4-weeks blocked in, they'll simply see "TEST". I'm thinking of leaving the choice of test (1rm or amrap) up to the lifter (which is total bullshit from a "research perspective" but the more lifter-friendly way to do things, I think.

Actually, it might not be that bad. If the lifter estimated their max going in to the program with an amrap then testing with an amrap would be consistent. I know my estimates coming in weren't perfect, so there's noise there (especially on DL with switching to Sumo) but hopefully it's not enough noise to mask a trend. I haven't gone for 1RMs on anything for a while, so I'm currently leaning that way.

On a different note. What kind of feedback do you want us to enter on the sessions? I logged my lifts but didn't know if I was supposed to include any commentary on lifts individually and/or the session overall.

[damufunman]

Now we're just getting into the "throw equations at it and see what fits what we think it should look like."
Sidenote: also why I don't like the Wilks factor calc.
ETA: @cgeorg Ok, it seems this was done jokingly...

It does seem that RPE per set may factor in as well, or at least something to factor in fatigue, maybe bar speed. To take into account the 5x3 vs 3x5 effect.

Actually, it might not be that bad. If the lifter estimated their max going in to the program with an amrap then testing with an amrap would be consistent. I know my estimates coming in weren't perfect, so there's noise there (especially on DL with switching to Sumo) but hopefully it's not enough noise to mask a trend. I haven't gone for 1RMs on anything for a while, so I'm currently leaning that way.

On a different note. What kind of feedback do you want us to enter on the sessions? I logged my lifts but didn't know if I was supposed to include any commentary on lifts individually and/or the session overall.

What about last squat 1RM was low bar and I've been doing high bar?

[Hanley]

but I'm wondering if they pressing can tolerate more.

Careful what you wonder about publicly. I added a set of 3 (so 6x3), then proposed another 2x2 at 80% (which is only a green-light based on real-time feels). You've got at least 72 hours recovery, so no huge whiff if the stress-dose exceeds 48 hours.

[Hanley]

On a different note. What kind of feedback do you want us to enter on the sessions? I logged my lifts but didn't know if I was supposed to include any commentary on lifts individually and/or the session overall.

If you RPE, you could give a range (say@6-7, or last set @8).

A purely qualitative summary is also valuable ("I was smoked", "This felt like warmup", "you call this a fucking workout?")/

[Hanley]

Can you give me 48-hour recoverable volume equivalents, in total reps, for 65%, 70%, 75%, 80%, 85%, and 90%?

I copied these from a thread in Egypt. I think they were starting points.

30 reps @70%
25 reps @75%
15ish @80%
5-8 @85%

I remember those. Doing the "non-traditional" approach to intraset fatigue (6x3 vs 3x6), I think I'd increase those numbers by at least 20%.

Except 85%. That stays.

[SpinyNorman]

What about last squat 1RM was low bar and I've been doing high bar?

Looks like you'll just have to run more than one cycle.

[SpinyNorman]

On a different note. What kind of feedback do you want us to enter on the sessions? I logged my lifts but didn't know if I was supposed to include any commentary on lifts individually and/or the session overall.

If you RPE, you could give a range (say@6-7, or last set @8).

A purely qualitative summary is also valuable ("I was smoked", "This felt like warmup", "you call this a fucking workout?")/

I haven't done RPE enough to be confident in it. I can probably get close when they get near @8, but I guess if it's not @8 then it's probably somewhere @6-7.

I was definitely not smoked, but getting that much work done in a reasonable amount of time was not a walk in the park either. I'll go back in and add some comments to the results.

[damufunman]

What about last squat 1RM was low bar and I've been doing high bar?

Looks like you'll just have to run more than one cycle.

Well, I still have about 3 hours to decide if I want to go back to low bar for a cycle (or three).
Actually, might be good to see where I can get back up to at the end since I'm certain low bar has detrained quite a lot due to Life things (baby) and having done high bar for the past 5ish weeks. Or high bar has helped. ¯\_(ツ)_/¯

[JonA]

but I think that's because I was keeping my rest intervals really short, like 2~ minutes, and didn't rest longer than it took to put the weight on the bar and redo my wrist wraps between warmups.

I kept rest pretty minimal as well. 60s for bench/dead, 90s for squat, as it felt pretty easy. I think I have been vastly over programming volume/intensity previously, or maybe my current e1RM are underestimated due to accumulated fatigue from my crap attempts to program myself.

[AaronM]

but I think that's because I was keeping my rest intervals really short, like 2~ minutes, and didn't rest longer than it took to put the weight on the bar and redo my wrist wraps between warmups.

I kept rest pretty minimal as well. 60s for bench/dead, 90s for squat, as it felt pretty easy. I think I have been vastly over programming volume/intensity previously, or maybe my current e1RM are underestimated due to accumulated fatigue from my crap attempts to program myself.

90s for bench, 4m for squat (they suck at any %), 2m for deads. I could've dropped the rest time on bench and deads and been fine I think.

[SpinyNorman]

What about last squat 1RM was low bar and I've been doing high bar?

Looks like you'll just have to run more than one cycle.

Well, I still have about 3 hours to decide if I want to go back to low bar for a cycle (or three).
Actually, might be good to see where I can get back up to at the end since I'm certain low bar has detrained quite a lot due to Life things (baby) and having done high bar for the past 5ish weeks. Or high bar has helped. ¯\_(ツ)_/¯

Did you switch to high bar just for something different to do?

I haven't even attempted a high bar squat in a long time. I'm sure it would feel strange now.

[SpinyNorman]

but I think that's because I was keeping my rest intervals really short, like 2~ minutes, and didn't rest longer than it took to put the weight on the bar and redo my wrist wraps between warmups.

I kept rest pretty minimal as well. 60s for bench/dead, 90s for squat, as it felt pretty easy. I think I have been vastly over programming volume/intensity previously, or maybe my current e1RM are underestimated due to accumulated fatigue from my crap attempts to program myself.

Wow, that's really short rest.

I started at 2.5 minutes on bench and waited 3 by the last set or two.
Squats were 3 minutes.
DL was 2.5 - 3 minutes, but the weight on that was fairly light.

[cgeorg]

Now we're just getting into the "throw equations at it and see what fits what we think it should look like."
Sidenote: also why I don't like the Wilks factor calc.
ETA: @cgeorg Ok, it seems this was done jokingly...

It does seem that RPE per set may factor in as well, or at least something to factor in fatigue, maybe bar speed. To take into account the 5x3 vs 3x5 effect.

Honestly I was just showing my work as I thought it out. I do think the last calculation is solid. Thinking about my feels under bar and the bar speed curve, there's big difference in the effect an @8, @9, or @10 effort have. Thinking about my feels under the bar, 6 reps @8 is less fatiguing than 3 reps @8. The difference in fatigue from adding extra sets is not that high, compared to those other factors. So a low x value drops the significance of set count in the first term, y is chosen so that an @10 is considered almost 3 fatiguing as taxing as an @6, and twice as taxing as an @7 in the second term, and the third term with a neutral z gives you a slight scaling of 3@8 being more fatiguing than 6@8.

Thinking about it a little more, the second term might make more sense as something like 1.2^(2*RPE) @10=40, @9=26, @8=18, @7=12, @6=9.

[PatrickDB]

Here is an observation. I don't claim I can produce a good fatigue metric. But I claim I can produce a better one than INOL.

For simplicity, I'm going to assume fatigue scales linearly in reps and that they're being done in the normal RPE 7-8 range, and ignore the question of what happens if you go to failure or do a bunch of sub-maximal work.

I propose a "Hanley normalized fatigue metric," or H for short, to be computed as:

[math]H = \text{reps}\times \left(\frac{100}{[100-\text{ intensity (in percent)}]}\right)^2 .[/math]


As a first check, we apply it to Hanley's table, using the larger value for 85%.

70%: H = 30*(100/30)^2 = 333.

75%: H = 25*(100/25)^2 = 400

80%: H = 15*(100/20)^2= 375

85%: H = 8*(100/15)^2 = 355.

This isn't a perfect fit, and the 70% value seems a little low. But in the context of the DUP program there's a little extra volume to compensate.

Bench hypertrophy day: 31*(100/30)^2 + 15*(100/40)^2 = 438

Bench strength day: 15*(100/20)^2= 375

If we add another set of 3, that's 450, which is right on the money with respect to hypertrophy day. The extra 2x2 might be too much. That puts us at 550, which is a lot! I might just skip the extra set of 3 and do the 2x2, to be honest, @Hanley. I'm not sure I can sustain 6 fast triples, but 2 more fast doubles probably works.

Bench power day: 5*(100/40)^2 + 4*(100/30)^2 + 2*(100/20)^2 + 2*(100/15)^2 = 215

This makes sense, since it's supposed to be a light recovery day. But there's some room here. I wonder if @mgil is right an the last single at 85% could be changed to a single at 90%. This adds 55 H. I wonder if that's affordable.

Just for fun, let's check the Barbell Medicine HLM template using this formula. The heavy bench day is 5 at 70%, 5 at 75%, then a total of 20 reps at 80%.

BBM heavy bench day: 5*(100/30)^2 + 5*(100/25)^2 + 20*(100/20)^2 = 635

Then there's some pressing on the light day, which I don't know how to count. It seems like a lot less stress than benching. Medium day is 12 total reps of 4@9 on close grip. That's probably about 85% of the close grip max, or say 80% of bench max. So another 12 reps gets us:

BBM medium day: 12*(100/20)^2 = 300.

Estimating the stress from the BBM light day pressing at about 200 (since that's about what our light day is), this gives about 1100 total H for the week for the BBM program. Hanley's program is about 1050 without the addition 2x2 on strength day, and 1150 with it.

It's interesting that this crude formula gives about the same overall weekly stress from two different HLM/HLH type programs with radically different rep schemes.

Appendix on a dumb light day calculation for the BBM template: my press e1RM is 140, my bench 1e1RM is 195, and 80% of 140 is about 115, or 60% of my bench max. Let us use this value for our calculation, even though it's completely unjustified.

Light day: 5*(100/45)^2 + 5*(100/47)^2 + 20*(100/40)^2 = 172

This checks out against my ballpark estimate for stress of 200.

cc: @chrisd @Chebass88

[JonA]

It does seem that RPE per set may factor in as well, or at least something to factor in fatigue, maybe bar speed. To take into account the 5x3 vs 3x5 effect.

I tried scaling the INOL by [reps in set / estimated max reps @ that %] and it evens things out quite nicely.

Eg, if you are doing sets of 5 reps @ 70%, you'd scale the INOL by 5/12. If you are doing sets of 3 @ 85%, it would be scaled by 3/5.

Applying that scale factor to this weeks workout for bench gives:

H
15 / 30 * (5/12) = .21
16 / 30 * (4/12) = .18
.39 SINOL

P
5.0 / 40 * (5/16)=0.04
4.0 / 30 * (4/12.0)=0.04
2.0 / 20 * (2/7.0)= 0.03
2.0 / 15 * (1/5.0)=0.03
=0.14 SINOL

S
15 / 20 * (3/7.0) = 0.32
= 0.32 SINOL

If, for example, if it was 3x5@80% rather than 5x3, it would be 15/20 *5/7 = .53, nearly double, which seems pretty likely.