1
00:00:01,079 --> 00:00:06,329
So it's my pleasure today to be here,
talking about new measurements that Atlas

2
00:00:06,359 --> 00:00:12,209
has been working on very diligently of
event shapes and jetsam structure of

3
00:00:12,209 --> 00:00:18,179
circles in the wrenchy data set. So in
particular, I'll be talking about three

4
00:00:18,659 --> 00:00:22,649
new measurements today. One is brand new
for this conference that was just approved

5
00:00:23,339 --> 00:00:28,349
this week, which is a measurement of event
shapes that high energy scales and their

6
00:00:28,349 --> 00:00:32,219
entry datasets. And then I'll also be
presenting two very recent measurements of

7
00:00:32,219 --> 00:00:36,269
jet substructure one of soft trough
observables, which is sort of looking at

8
00:00:36,269 --> 00:00:40,019
the same observables that we've already
heard about from the least actually, the

9
00:00:40,019 --> 00:00:45,359
CG and RG soft dropcam chips, and then
another which is looking at the lunch at

10
00:00:45,359 --> 00:00:49,289
plan which has already actually been
introduced in the session. I just wanted

11
00:00:49,289 --> 00:00:51,899
to say at the start of my talk that all of
the results that I'm going to be showing

12
00:00:51,899 --> 00:00:56,489
are unfolded. We use an iterative vision
unfolding technique and Atlas for details

13
00:00:56,489 --> 00:01:00,239
about the way this is done for one
measure, mentor and other than that I

14
00:01:00,239 --> 00:01:04,739
would please refer you to the individual
publications. And I also wanted to say

15
00:01:04,739 --> 00:01:10,109
that all of our metric data are or will be
assumed available on EPA. And that we

16
00:01:10,109 --> 00:01:13,229
provide our analysis routines and rivets
and some of them are already in place,

17
00:01:13,229 --> 00:01:18,179
actually. So moving on to event shapes.
This is something that most people here

18
00:01:18,179 --> 00:01:21,689
are probably already very familiar with.
Then shapes are a family of observables,

19
00:01:21,689 --> 00:01:26,999
which are sensitive to event wide energy
flow. There's sort of a staple of GCD

20
00:01:26,999 --> 00:01:30,299
programs, they have a lot of different
applications from testing Monte Carlo

21
00:01:30,299 --> 00:01:32,249
generators to measuring alpha s,

22
00:01:33,600 --> 00:01:35,940
to even searching for physics beyond the
standard model.

23
00:01:37,200 --> 00:01:38,190
And there's been a lot of

24
00:01:38,970 --> 00:01:44,280
improvement both experimental and in terms
of Monte Carlo estimates. Since the last

25
00:01:44,280 --> 00:01:48,720
time Atlas measured event shapes and
Multijet events, which was in a data set

26
00:01:48,720 --> 00:01:53,310
that we collected years ago this
direction. So it goes without saying that

27
00:01:53,310 --> 00:01:57,630
we have much more data in the renter
dataset we did in the 2010 data set, which

28
00:01:57,630 --> 00:02:02,850
enables new multi differential
measurements to be made. We also have

29
00:02:02,850 --> 00:02:07,380
better objects. So jet reconstruction has
improved substantially. In the last 10

30
00:02:07,380 --> 00:02:11,490
years, we now have percent level
uncertainties on high PT jets. Here I'm

31
00:02:11,490 --> 00:02:16,710
showing our in situ calibration and its
associated uncertainty for our new

32
00:02:16,740 --> 00:02:20,250
Particle Flow jets which were used to
perform this measurements. For more

33
00:02:20,250 --> 00:02:23,550
details on check calibration and
performance analysis, I would refer you to

34
00:02:23,550 --> 00:02:25,710
Peters talk from earlier this week.

35
00:02:27,570 --> 00:02:29,730
So we look at several different events
shapes,

36
00:02:30,810 --> 00:02:34,470
I won't have time to talk about all of
them in details. thrust is one that you're

37
00:02:34,470 --> 00:02:39,780
probably familiar with, which interpolates
between digests like events and isotropic

38
00:02:39,780 --> 00:02:45,360
events. Other event shapes are defined in
terms of the linear linearized scarcity

39
00:02:45,360 --> 00:02:48,870
tensor, which is shown sort of the big
equation in the middle of the slide.

40
00:02:49,920 --> 00:02:50,550
Or

41
00:02:50,700 --> 00:02:55,470
so that would be the sphericity and the
clarity, which tells you how sort of

42
00:02:55,470 --> 00:02:56,910
isotropic or planar events that

43
00:02:57,899 --> 00:02:59,789
if you take its transverse projection

44
00:03:00,810 --> 00:03:04,980
Then you can also look at other event
shapes like the transverse sphericity, or

45
00:03:04,980 --> 00:03:10,140
A, C and D, which are event shapes related
to three unfortunate events. To illustrate

46
00:03:10,140 --> 00:03:13,530
this, you could sort of think about the
two events that I show on the right side

47
00:03:13,530 --> 00:03:18,570
of the slide here. With this a five jet
event with small thrust and small

48
00:03:18,570 --> 00:03:23,190
transfers. sphericity is that this is a
very digest like a pencil like event, or

49
00:03:23,190 --> 00:03:27,060
the three jet event below it, which has
large thrust, large transfer sphericity.

50
00:03:27,060 --> 00:03:31,140
So this is a more isotropic distribution
of energy. All of these measurements are

51
00:03:31,140 --> 00:03:36,450
triple differential in terms of the ht, so
the energy scale, the jet multiplicity and

52
00:03:36,450 --> 00:03:39,150
the event chips themselves. So here

53
00:03:39,150 --> 00:03:40,080
I'm showing one,

54
00:03:41,010 --> 00:03:45,420
which is the rarity, so this interpolates
between planar and less planar, more

55
00:03:45,420 --> 00:03:53,040
isotropic events in the ht band from one
to 1.5 T. In the middle of the slide, you

56
00:03:53,040 --> 00:03:58,020
can see these distributions. These are the
ratios for different numbers of jets

57
00:03:58,830 --> 00:04:02,520
between the measured data And several
different Monte Carlo generators which

58
00:04:02,520 --> 00:04:07,530
we're comparing to. So you kind of have to
stare at these plots for a long time. But

59
00:04:07,950 --> 00:04:11,970
something starts to become apparent.
There's a lot of information here. So in

60
00:04:11,970 --> 00:04:16,020
particular, this event shape seems to be
sensitive to the potential or algorithm

61
00:04:16,020 --> 00:04:21,180
that's being studied. So here you can see
a comparison of herb extended and run with

62
00:04:21,180 --> 00:04:24,480
either an Angular ordered or a dipole
shower and you can see for instance, in

63
00:04:24,480 --> 00:04:29,640
the three jet in the purple in the green
lines grow quite far apart from each other

64
00:04:29,640 --> 00:04:34,650
as you move towards less planar
arrangements of energy. pithy attempts to

65
00:04:34,650 --> 00:04:40,350
describe the planar events quite well.
Well, AMC anello integrates with 50 after

66
00:04:40,350 --> 00:04:46,320
the proton shower also performs quite well
in the low jet fins here and that low ht

67
00:04:46,470 --> 00:04:50,490
This measure is actually dominated by
modeling uncertainties, which is sort of

68
00:04:50,490 --> 00:04:51,450
interesting to see.

69
00:04:53,279 --> 00:04:54,959
If you look at other

70
00:04:55,379 --> 00:05:00,989
events shape, so here, this is the
transfers trust. So penciled like events

71
00:05:00,989 --> 00:05:04,019
are on the left side of the plot isotropic
ones on the right side,

72
00:05:05,069 --> 00:05:07,379
you can see some similar

73
00:05:07,499 --> 00:05:08,099
trends.

74
00:05:08,789 --> 00:05:12,059
But the systematic uncertainties in
particular here are very different. So

75
00:05:12,059 --> 00:05:14,819
here we're dominated by the jet
uncertainties, which is maybe what you

76
00:05:14,819 --> 00:05:19,289
would expect for a measurement of event
shapes. But this only happens in the most

77
00:05:19,289 --> 00:05:23,279
extreme regions that face space that are
probed by this measurements. In terms of

78
00:05:23,699 --> 00:05:27,119
observations about the measurement of
transfers, trust pythia tends to do quite

79
00:05:27,119 --> 00:05:31,769
well in direct light topology, which is
what you would probably expect. The

80
00:05:31,769 --> 00:05:36,449
relative cross sections as a function of
the jet multiplicity tend to vary quite a

81
00:05:36,449 --> 00:05:40,679
bit for the different generators. So you
can see that her weight overestimates how

82
00:05:40,679 --> 00:05:44,759
many events there are in the three jet
then and underestimates in the six jet

83
00:05:44,759 --> 00:05:51,269
spin and Sherpa sort of shows the opposite
trend here. I'm going to move on to talk

84
00:05:51,269 --> 00:05:55,859
about our substructure measurements,
starting with our mission and stop shop

85
00:05:55,859 --> 00:06:00,779
observables. So before we can actually
compare jet substructure theory we have to

86
00:06:00,779 --> 00:06:04,979
do a little bit of work. Soft drop has
already been introduced. So I won't spend

87
00:06:04,979 --> 00:06:08,819
a lot of time talking about the algorithm
itself. But just a little refresher for

88
00:06:08,819 --> 00:06:12,179
those connected. This is an auntie Katie
jet.

89
00:06:12,600 --> 00:06:15,570
You can tell because the heart energy
clusters have been

90
00:06:15,600 --> 00:06:18,840
sort of associated with each other first
and then stuff things have been clustered

91
00:06:18,840 --> 00:06:23,970
into those. If we want to perform subtract
them into this chat, what we have to do

92
00:06:23,970 --> 00:06:28,020
first is re cluster the constituents of
the jet using the Cambridge jaken

93
00:06:28,110 --> 00:06:33,810
algorithm, which gives you an Angular
order. And then we can apply the soft drop

94
00:06:33,810 --> 00:06:37,050
algorithm moving from the widest angles
ambition, so this would be following the

95
00:06:37,050 --> 00:06:42,450
nodes on the graph moving from the left to
the right. at each step, we can check

96
00:06:42,450 --> 00:06:46,920
whether the softer subjects satisfies the
soft drop condition, which you can see

97
00:06:46,920 --> 00:06:51,660
pictured on the slide. This depends on two
parameters set cap and beta. Beta in

98
00:06:51,660 --> 00:06:54,720
particular is important for the
measurement that we'll be presenting.

99
00:06:55,560 --> 00:07:00,270
higher values of beta mean that we grew
less, so beta zero is destroyed. just

100
00:07:00,270 --> 00:07:05,160
screaming that we'll be talking about.
When a smiting satisfies a softer

101
00:07:05,160 --> 00:07:09,030
condition, we stopped. And this is the
grim jet that we can actually measure the

102
00:07:09,030 --> 00:07:14,760
properties of and compare them to
analytical predictions. So we tried to

103
00:07:14,760 --> 00:07:20,130
fully characterize this splitting within
the jet, where the originating part time

104
00:07:20,160 --> 00:07:24,060
emits a large fraction of its energy which
is enough to stop this after after main

105
00:07:24,060 --> 00:07:28,920
procedure. So, we mentioned three
observables, the massive the jet. This is

106
00:07:28,920 --> 00:07:32,910
the first substructure of circle which was
available with a theoretical position

107
00:07:32,940 --> 00:07:36,450
above leading loads and resubmission. And
it's the one where the most precise

108
00:07:36,450 --> 00:07:41,070
calculations are available. We also
measured the opening angle or Archie. This

109
00:07:41,070 --> 00:07:46,080
is a variable where new predictions that
next deleting log were available. And we

110
00:07:46,110 --> 00:07:50,730
also look at the PT balance of the heart
splittings, the CG observable that was

111
00:07:50,730 --> 00:07:57,360
introduced in the previous talk, which is
also interesting to look at. So we measure

112
00:07:57,360 --> 00:08:00,150
these quantities in many different ways. I
won't be able to talk about all The

113
00:08:00,150 --> 00:08:06,510
results in this paper today, but we study
performance of measurements made up of

114
00:08:06,510 --> 00:08:11,550
clusters that the calorimeter are charged
on a track jets. They're all done teaching

115
00:08:11,550 --> 00:08:17,310
differentials starting at 300 gV with R
equals 0.6 chips and going up. And we

116
00:08:17,310 --> 00:08:22,560
studied multiple different stuff drop
settings as well. So here are a comparison

117
00:08:22,590 --> 00:08:27,780
of the three variables that we're looking
at. between data and Monte Carlo

118
00:08:27,780 --> 00:08:34,590
generators, you can see performance,
broadly speaking, is quite good. Although

119
00:08:34,740 --> 00:08:38,700
agreement between the different generators
tends to degrade a bit when you enter

120
00:08:38,700 --> 00:08:42,480
regions to face space where there are
larger non perturbative effects. So places

121
00:08:42,480 --> 00:08:47,700
like a spa uses a mess. But we have very
asymmetric splittings in this energy

122
00:08:47,700 --> 00:08:54,270
distribution. The agreement here could be
a little bit better. Moving on to the next

123
00:08:54,270 --> 00:09:00,720
slides. Here I compare the data to the
analytical justice About the analytical

124
00:09:00,720 --> 00:09:04,470
calculations that we show in the paper,
the masses on the left, you can see three

125
00:09:04,470 --> 00:09:07,710
different predictions are compared to the
data. And the important region of this

126
00:09:07,710 --> 00:09:13,020
plot is really the central region where
I've highlighted with a box. So this area

127
00:09:13,020 --> 00:09:17,010
is dominated by resubmission. And it's
where the calculations are sort of

128
00:09:17,010 --> 00:09:21,660
designed to predict most accurately at
lower values at the mass number attribute

129
00:09:21,660 --> 00:09:26,670
as effects become large and higher by
instant master sort of dominated by fixed

130
00:09:26,670 --> 00:09:31,620
earth fix. The equipment in resubmission
region is quite good. And the precision

131
00:09:31,650 --> 00:09:37,260
here is also quite impressive, sort of
comparable between the theory and the

132
00:09:37,260 --> 00:09:45,150
data. On the right side of the of the
slide, I show you two values of RGB. So

133
00:09:45,150 --> 00:09:49,290
this is the green jet radius or the
opening angle for different values of

134
00:09:49,290 --> 00:09:53,310
beta. So there's more grooving on the top
and less turning on the bottom and you can

135
00:09:53,310 --> 00:09:59,160
see that as we proved less, non
conservative effects become larger and the

136
00:09:59,160 --> 00:10:01,710
agreement between The data in the theory
sort of degrades here.

137
00:10:05,309 --> 00:10:06,149
Now moving on,

138
00:10:07,200 --> 00:10:10,740
we can talk about more than just the
hardest lifting with the ninja. And in

139
00:10:10,740 --> 00:10:15,090
fact, another measurement that we've done
studies the entire jet cluster in history.

140
00:10:15,630 --> 00:10:20,100
So you might want to think of it a jet
here as a cloud of soft glue in admissions

141
00:10:20,100 --> 00:10:24,210
around a hardcore, which would be the
originating parts on, you can parameterize

142
00:10:24,210 --> 00:10:28,140
those submissions in terms of their
relative angle and energy to correct the

143
00:10:28,140 --> 00:10:33,720
check. We get this in practice by looking
again at our camper, chicken cluster jet

144
00:10:33,780 --> 00:10:37,260
history. And each node in the jet looking
at the

145
00:10:38,100 --> 00:10:39,150
software subject.

146
00:10:41,070 --> 00:10:46,650
Looking at that subjects relative angle
and energy to the jet core and taking this

147
00:10:46,650 --> 00:10:53,430
as approximately the mission, you know, in
our case, and using this to put a point in

148
00:10:53,430 --> 00:10:58,680
the lunch at plane so this is a face
space, which is sort of spinning these two

149
00:10:58,680 --> 00:11:04,260
variables, the relative energy In the
angle of the admissions within the jet. So

150
00:11:04,260 --> 00:11:08,790
you can see for one jet, what the plane
looks like on the right side of the slide.

151
00:11:08,820 --> 00:11:14,640
So this is a set of points in this 2d
space. And here we show it at a critical

152
00:11:14,640 --> 00:11:19,650
level and detector level. So each node in
the graph in the middle of the slide

153
00:11:19,650 --> 00:11:26,250
corresponds to one point as we filled the
template. Now this is what it looks like

154
00:11:26,250 --> 00:11:31,890
for one jet. If you do this for 30 million
jets in the red two data set that high PT,

155
00:11:31,920 --> 00:11:37,470
this is what you end up with. So this
picture is actually quite interesting if

156
00:11:37,470 --> 00:11:41,010
you stare at it and think about GCP for a
while, so I'll walk you through some of

157
00:11:41,010 --> 00:11:45,360
the observations that you can make. So the
way this works is that wide angle then

158
00:11:45,360 --> 00:11:49,170
even splitting so towards the bottom left
corner, as you move from left to right,

159
00:11:49,410 --> 00:11:53,370
you look at more collinear splittings
within the jet, and as you move from the

160
00:11:53,370 --> 00:11:56,130
bottom to the top, you look at softer
splittings within the gym.

161
00:11:57,809 --> 00:11:59,999
So the bottom left corner is

162
00:12:01,380 --> 00:12:06,570
Cool, this region gets populated
relatively uniformly by nice perturbative

163
00:12:06,570 --> 00:12:11,010
qc D emissions, which is sort of what you
would expect if you think about looking at

164
00:12:11,010 --> 00:12:17,970
the lead plan in terms of the leading luck
prediction, and it's nicely protected from

165
00:12:18,030 --> 00:12:22,620
non perturbative effects which enter the
rest of the picture. The feature in the

166
00:12:22,620 --> 00:12:26,100
center is actually caused by
hybridization. In this region, the

167
00:12:26,100 --> 00:12:33,780
probability of a jet emitting some soft
radiation increases. This is mostly due to

168
00:12:33,840 --> 00:12:39,660
the running of alpha s, and the fact that
hadron ization happens. This region is

169
00:12:39,660 --> 00:12:43,470
particularly sensitive to the choice of
hybridization models and sort of soft

170
00:12:43,470 --> 00:12:51,180
dynamics within the region beyond this
feature, a strongly suppressing submission

171
00:12:51,180 --> 00:12:56,340
so there's very little in the upper right
corner of the plot. If you take a profile

172
00:12:56,340 --> 00:12:59,790
through the lunch lens, so here I'm
showing you a plot on the right side of

173
00:12:59,790 --> 00:13:05,460
the screen. lives where we've taken a
slice horizontally, then you can actually

174
00:13:05,460 --> 00:13:11,220
show some other very nice features of this
observable. Here you can see that

175
00:13:11,400 --> 00:13:17,430
potential our models tend to differ quite
a lot on the right side of the ledger plan

176
00:13:17,610 --> 00:13:21,960
and Hedren ization models tend to differ
quite a lot in terms of their prediction

177
00:13:21,960 --> 00:13:24,390
on the right side. So these are the
differences between

178
00:13:25,650 --> 00:13:28,050
two photon shower algorithms on the left.

179
00:13:29,789 --> 00:13:33,899
They open and close Greenmarkets and 200
ization models on the right the open and

180
00:13:33,899 --> 00:13:39,449
close orange methods whatever the green
markers come together, where the pattern

181
00:13:39,449 --> 00:13:40,349
ization models

182
00:13:41,159 --> 00:13:42,659
split apart and vice versa.

183
00:13:44,520 --> 00:13:48,540
So in this way, we can talk about how this
observable factor has its different

184
00:13:48,540 --> 00:13:54,630
physical effects. We achieve a typical
position of about 10% as a function of the

185
00:13:54,630 --> 00:14:01,350
legend claim. This uncertainty is by and
large dominated by MC modeling effects. No

186
00:14:01,350 --> 00:14:06,120
single simulation out of the six that we
compare to the unfolded data really

187
00:14:06,120 --> 00:14:06,600
describes

188
00:14:06,600 --> 00:14:08,370
the entire metric region perfectly.

189
00:14:11,400 --> 00:14:13,290
Okay, and I've already talked about the
factorization.

190
00:14:15,179 --> 00:14:18,629
to really emphasize the way that the lunch
at play and factor is, is different

191
00:14:18,629 --> 00:14:25,559
physical effects, I have these two extra
slides or two extra plots on slide 21. So,

192
00:14:25,889 --> 00:14:29,729
on the left, you can see the actual 2d
ratio of the lens land for the different

193
00:14:29,729 --> 00:14:33,599
potential our algorithms that we have in
Hurley. And you can see this region of

194
00:14:33,599 --> 00:14:38,789
sensitivity in the bottom left corner
where we expect these sort of harder wider

195
00:14:38,789 --> 00:14:43,469
angle emissions to be. And on the right
slide, on the right side of the slide, I

196
00:14:43,469 --> 00:14:49,439
show the ratio of the two different Sherpa
models and where the heparin ization

197
00:14:49,439 --> 00:14:52,919
effects different than most disappointment
to play. And you can see that these are

198
00:14:53,009 --> 00:14:55,469
basically completely different regions
which is really nice.

199
00:14:58,110 --> 00:15:02,130
Okay, so chicken Sorry.

200
00:15:04,740 --> 00:15:09,900
Hello, all fine, please go ahead. I remind
you, but you have one minute left.

201
00:15:11,070 --> 00:15:17,850
Ah, perfect. Okay, great. Yes. Okay, so to
wrap up, there's a rich program of jets of

202
00:15:17,850 --> 00:15:22,770
structure measurements being pursued by
all four Elysee collaborations in Russia,

203
00:15:23,130 --> 00:15:27,540
I think this is very clear, actually from
watching a lot of the other talks at this

204
00:15:27,540 --> 00:15:31,530
conference this week. And I think that
Atlas is contributing very novel analysis

205
00:15:31,530 --> 00:15:39,210
to these efforts, which is very nice to
see the larger entry data set, and the

206
00:15:39,210 --> 00:15:43,890
really, really nice Atlas, Jet
reconstruction performance, make this time

207
00:15:43,890 --> 00:15:47,550
periods, sort of the perfect time to
revisit some of these old bread and butter

208
00:15:47,550 --> 00:15:53,160
measurements like event shapes. We're
starting to do that as well. In many

209
00:15:53,160 --> 00:15:57,810
places, I've mentioned this a couple of
times, we're limited by large Monte Carlo

210
00:15:57,810 --> 00:16:02,190
modeling related uncertainties. So I also
wanted to mention As I wrap up that we are

211
00:16:02,220 --> 00:16:05,700
quite interested in cooperating with the
wider community in order to try to improve

212
00:16:05,700 --> 00:16:09,540
the position of our simulation, because
this is limiting the position of our

213
00:16:09,540 --> 00:16:13,530
measurements. So we're interested in
knowing what the most helpful data we

214
00:16:13,530 --> 00:16:16,200
could provide in order to facilitate that
would be

215
00:16:17,760 --> 00:16:19,140
okay, thanks for your attention.

216
00:16:21,539 --> 00:16:27,629
Thank you much for this very interesting
results. So now the wider community please

217
00:16:27,629 --> 00:16:28,919
ask questions.

218
00:16:39,690 --> 00:16:41,760
Can't believe there are no questions.

219
00:16:48,210 --> 00:16:55,230
I okay. So let me ask I had one question,
but it's more of an outsider question to

220
00:16:55,230 --> 00:17:03,510
your slide 14. Okay.

221
00:17:09,419 --> 00:17:16,769
Yeah, exactly. So I'm just wondering,
what's the difference between the purple

222
00:17:16,829 --> 00:17:24,029
on the left hand plot between the purple
and the red because I suppose, and Ll is

223
00:17:24,029 --> 00:17:29,369
also in the purple also includes leading
others. So the only difference is the non

224
00:17:29,369 --> 00:17:31,919
perturbative effects as opposed.

225
00:17:41,910 --> 00:17:45,870
Okay, I'll try again. I'll turn it back.
So the different predictions are provided

226
00:17:45,870 --> 00:17:49,440
by very different groups actually. So
they're using different methodologies, the

227
00:17:49,440 --> 00:17:54,240
NLL here. I think it's actually a
prediction for inclusive jets. So it is

228
00:17:54,330 --> 00:18:00,450
an nl nll it is exactly what it says here.
You can

229
00:18:00,450 --> 00:18:04,770
see also on the right that this is an nll
prediction for for RG

230
00:18:06,270 --> 00:18:09,150
which is also a prediction done for
inclusive jets and then we

231
00:18:09,420 --> 00:18:12,210
compare it to this measurement which is
actually done in digests.

232
00:18:13,140 --> 00:18:17,520
Whereas the red part does include a sort
of

233
00:18:18,450 --> 00:18:19,710
the Kickstarter part.

234
00:18:23,220 --> 00:18:28,140
I see. Okay, thanks. Yeah, I thought the
fixed other part that always always

235
00:18:28,140 --> 00:18:31,560
included but okay. Yes, thank you.

236
00:18:32,340 --> 00:18:35,460
Yeah, I'm not an expert, but this is my
understanding of the difference between

237
00:18:35,460 --> 00:18:36,810
these predictions.

238
00:18:40,710 --> 00:18:42,630
Any other questions?

239
00:18:51,120 --> 00:18:51,840
No.

240
00:18:53,460 --> 00:18:56,880
Then Thank you Mark. Again.